Research ArticleAn Improved Shuffled Frog Leaping Algorithm forAssembly Sequence Planning of Remote Handling Maintenancein Radioactive Environment

Jianwen Guo1 Hong Tang2 Zhenzhong Sun1 Song Wang3 Xuejun Jia4

Haibin Chen1 and Zhicong Zhang1

1School of Mechanical Engineering Dongguan University of Technology Dongguan 523808 China2Dongguan Neutron Science Center Dongguan 523890 China3College of Mechanical amp Automotive Engineering South China University of Technology Guangzhou 510641 China4Institute of High Energy Physics Chinese Academy of Sciences Beijing 100049 China

Correspondence should be addressed to Jianwen Guo guojwdguteducn

Received 2 October 2014 Revised 13 January 2015 Accepted 28 January 2015

Academic Editor Massimo Zucchetti

Copyright copy 2015 Jianwen Guo et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Assembly sequence planning (ASP) of remote handling maintenance in radioactive environment is a combinatorial optimizationproblem This study proposes an improved shuffled frog leaping algorithm (SFLA) for the combinatorial optimization problem ofASP An ASP experiment is conducted to verify the feasibility and stability of the improved SFLA Simultaneously the improvedSFLA is compared with SFLA genetic algorithm particle swarm optimization and adaptive mutation particle swarm optimizationin terms of efficiency and capability of locating the best global assembly sequence Experiment results show that the proposedalgorithm exhibits outstanding performance in solving theASP problemThe application of the proposed algorithm should increasethe level of ASP in a radioactive environment

1 Introduction

Radioactive installation (eg nuclear power plants andhigh-energy physics research institutes) generally has thecharacteristics of complicated structures high speed andheavy loads The installations themselves and their workingenvironments are radioactive The aforementioned charac-teristics may cause failures of key equipment in radioactiveinstallation which seriously affect the lifetime of radioactiveinstallation

Maintenance refers to restoring aging or faulty equip-ment parts to a satisfactory operating condition It includesinspection testing diagnosis disassembly assembly clean-ing repair and replacement Key equipment of radioactiveinstallation that provides base functions must be maintainedduring installation lifetime [1ndash3] In radioactive installationmost maintenance activities are conducted in a radioactiveenvironment that is unsuitable for humans in such casesremote handling maintenance (RHM) is necessary [4] RHM

enables a person to manually handle work without beingphysically present at a work site through a manipulator or arobot [5]

Radioactive equipment has a complex structure thatcauses difficulty in maintenance operations within a radioac-tive installationMaintenance procedures must be planned inadvance to ensure reliability and security of RHM[6] Remotehandling maintenance planning (RHMP) predetermine themaintenance procedures of radioactive equipment duringthe design of radioactive installation [7] Assembly sequenceplanning (ASP) determines the order in which each part andsubassembly must be inserted into an incrementally expand-ing subassembly that eventually leads to a final assembly [8]Assembly operation is part of RHM procedure Thus ASP isconsidered as a subdomain of RHMP

In RHM ASP provides an optimal sequence to replaceaging or faulty parts under certain constraint conditions (egtime cost and reliability) Complex radioactive equipmentwith a large number of parts has several feasible sequences

Hindawi Publishing CorporationScience and Technology of Nuclear InstallationsVolume 2015 Article ID 516470 14 pageshttpdxdoiorg1011552015516470

2 Science and Technology of Nuclear Installations

for assembly Finding an optimal assembly sequence to satisfytime cost and reliability requirements is a combinatorialproblem the complexity of this problem is proportional tothe number of equipment parts The number of feasibleRHM sequences increases with equipment complexity [9] Alarge number of equipment parts results in a combinationexplosion and the optimal solution is omitted ASP is thenshown to be NP-complete [10]

Optimization techniques based on principles inspired bynatural systems have been proposed over the past decades tosolve the combinatorial explosion problem [11] The shuffledfrog leaping algorithm (SFLA) is a recent metaheuristicoptimization algorithm that is inspired by the memeticevolution of a group of frogs when seeking food SFLAinvolves a set of frogs that cooperate to achieve a unifiedbehavior for the entire system which produces a robustsystem that can find high-quality solutions to problemswith large search spaces [12] SFLA exhibits global searchingcapability rapid convergence and strong robustness It hasbeen successfully applied to several fields However SFLAis suitable for continuous optimization [13] SFLA should beimproved when applied to ASP which is a discrete search andoptimization problem

In this study an improved SFLA is presented to solvethe ASP problem for RHM in radioactive environment EachSFLA operation is redefined In particular a swap operatorand a swap sequence are introduced and the local searchstrategy is designed to directly search the discrete domainA diversity control strategy based on genetic algorithm (GA)is proposed to improve the search for global optimal solu-tions An ASP experiment show that the algorithm exhibitsoutstanding performance in solving the ASP problem

The remainder of the paper is organized as followsSection 2 introduces related works Section 3 describes SFLASection 4 states the ASP problem Section 5 discusses theimproved SFLA for ASP Section 6 describes the experimentsand the analyses Finally Section 7 provides the conclusionsof the study

2 Related Studies

21 RHM Maintenance preserves or restores a system orfacility to its desired state The following problems shouldbe considered for maintenance in radioactive installation (1)safety of the maintenance worker in cases where humanscannot gain access because of the high radiation dose rate (2)feasibility of maintenance work in cases where humans hasdifficulty working with equipment because of certain condi-tions (eg small spaces and narrow gaps) and (3) reliabilityofmaintenance work in cases where harsh environments andheavy workloads cause human errors

RHM is applied to solve the aforementioned problems[3ndash5] RHM mainly repairs fault parts that cause equipmentto stop working in a radioactive environment Operationsof RHN mainly include replacement and disposal workswhich are remotely handled by using power and master-slavemanipulators [14]The following observations aremade(1) RHM differs from conventional equipment maintenance

because it employs a robot or a remote operation tool in a hotcell instead of a human (2) The robot must be teleoperatedfully controlled or supervised by a human outside the hot cellbecause the majority of RHM tasks require the intuition andintelligence of a human [5] (3) A human does not need tobe physically present at the work site to conduct maintenancework

RHM will be applied to radioactive installation suchas the China Fusion Engineering Test Reactor [15] theInternational Thermonuclear Experimental Reactor (ITER)[6] and the European Organization for Nuclear Research[16]

22 Robots in RHM Several robots have been developed forRHM Takeda et al [17] designed three kinds of robots thatcan transport different parts in a radioactive environmentThe French Atomic Energy Agency Interactive RoboticsLaboratory developed an industrial robot system for a nuclearspent fuel reprocessing plant [18] The robot which usesRX170 as a slave arm and a control platform called TAO2000V2 supports a master-slave operation with a force feedbackand tolerates radiation up to a 10 kGy integrated dose [19]Sanders [20] developed a remote handling system with ldquomanin the looprdquo approach that provides the remote robot operatorfor the Joint European Torus Vale et al [21] developedan autonomous mobile robot for ITER Terada et al [22]designed and developed a pick-and-place work robot tocope with the module placement for the semiconductortracker barrel assembly The robot can place modules with amechanical precision of over 25 120583m Lee et al [23] developeda bridge transported servo-manipulator system to overcomethe limited workspace of conventional mechanical master-slave manipulators in a hot cell Lee et al [24] developeda cable-driven dual arm master-slave servo-manipulator forthe pyroprocess research facility

23 RHMP In radioactive installation RHMP predeter-mines the maintenance process during the stages of radioac-tive installation design Traditional RHM is mainly per-formed through an empirical design and physical verificationby experiment platform It is unsuitable for complex structureinstallation and can be laborious and ineffective The experi-ment platform results in high costs and a long cycle

The development of computer artificial intelligence andsimulation technologies allows the application of virtualmaintenance planning to RHM [25] Takeda et al [26]developed a virtual reality simulator to support the Banketsimulation of theRHMprocess of ITERHeemskerk et al [27]studied the simulation process dynamics based on the ITERRHM simulator Geng et al [28] developed a novel virtualmaintenance application for maintenance safety evaluationto provide recommendations on maintenance safety Esqueet al [29] completed a digital simulation model of the ITERseparator Shuff et al [30] developed a set of discrete eventsimulation tools for the remote operating process planning ofthe ITER hot cabinet Robbins et al [31] achieved a real-timeand visualized track of remote operating process planning byusing virtual reality and an intelligent database Park et al [32

Science and Technology of Nuclear Installations 3

33] studied visualization and simulation of a nuclear facilitydisassembling process and established an RHM system

In the aforementioned studies maintenance sequencewas generally obtained in an exploratory manner because ofthe lack of a guided optimized maintenance sequence Suchsequencemay be inefficient and the optimal solutionmight beomitted without intelligent support Introducing intelligentplanning technologies to ASP is essential to further enhancevirtual maintenance planning [7]

24 Intelligent ASP for Complex Products Assembling prod-ucts (by humans or by robots) is the act of combiningparts of equipment ASP obtains the order for each part andsubassembly which is then inserted into an incrementallyexpanding subassembly that eventually leads to a final assem-bly ASP is a combinatorial problem in complex productsSolving this problem with human involvement is difficultand impractical because of the combinatorial explosion issueResearch in intelligence assembly sequencing has rapidlyincreased in recent years The intelligence ASP problem isregarded as a discrete search and optimization problemVarious artificial intelligence approaches have been proposedrecently including graph theory [34 35] subassembly detec-tion [36 37] motion planning [38] and evolution algorithm[11]

Evolution algorithms provide new solutions to vari-ous complex optimization problems by imitating the self-organization mechanism of natural biological communitiesand the adaptive ability of evolution Chen and Xiao [39]developed an enhancing artificial bee colony algorithmwith self-adaptive searching strategy and artificial immunenetwork operators for global optimization Xu et al [40]developed an improved genetic algorithm for distributionnetwork planning Chen and Ju [41] developed a novelartificial bee colony algorithm for solving the supply chainnetwork design under disruption scenarios Cheng et al[42] developed a metaheuristics for airport gate assignmentLorpunmance and Sap [43] developed an ant colony opti-mization for dynamic job scheduling in grid environmentEvolution algorithms such as GA [44ndash46] the ant colonyalgorithm [47 48] the particle swarm algorithm [49ndash52] andthe artificial bee colony algorithm [53] have been studiedrecently in ASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems and hasbeen found to be effective in searching for global solutionsRahimi-Vahed andMirzaei proposed a hybrid multiobjectiveSFLA for a mixed model assembly line sequencing prob-lem [54] and bicriteria permutation flow shop schedulingproblem [55] Li et al [56] proposed an effective SFLA fora multiobjective flexible job shop scheduling problems Fangand Wang [57] introduced an effective SFLA for a resource-constrained project scheduling problem Li et al [58] devel-oped a modified SFLA for continuous optimization Theaforementioned studies show that SFLA is a simple robustand fast algorithm for solving combinatorial optimizationproblems

Start

(3) Memetic evolution (by local search)

(4) Global shuffling process

Yes

(2) Grouping operation

(1) Population initialization and parameter initialization

Convergence criterions are satisfied

Output the optimal solution

End

No

Figure 1 SFLA procedure

3 SFLA

SFLA is a metaheuristic optimization method that identifiessolutions by simulating collaboration behaviors and inter-active information similar to a frog community searchingfor food in a natural environment This algorithm dividespopulation into several subpopulations and the evolutionof memes is driven by the exchange of global informationamong subpopulations and the local evolutionary searchwithin subpopulation

SFLA is described in detail as follows The populationconsists of several frogs and each frog is a solution to theproblem The population is divided into several subpopula-tions through a grouping operator to simulate frog groupingbehaviors Each subpopulation is called a memeplex Amemeplex is composed of frogs with the same meme thatperform local searches Each frog has its own idea but isalso influenced by other frogs in the same memeplex Frogsadjust their positions through memetic evolution After apredefined number of memetic evolutionary times frogs indifferent subpopulations exchange information through aglobal shuffling process The alternating memetic evolutionand global shuffling process make the frogs jump out of thelocal optimum and evolve toward the global optimum

The SFLAprocedure is illustrated in Figure 1Thedetailedprocedure is as follows

(1) Population initialization and parameter initializationSFLA initially creates a population of 119865 frogs as acertain solution amount The 119894th frog (ie the 119894thsolution to the problem) is represented as 119865

119894=

4 Science and Technology of Nuclear Installations

(1198911198941 1198911198942 119891

119894119904) where 119904 is the solution space dimen-

sion(2) Grouping operator a grouping operator separates119865 frogs into 119898 memeplex which contains 119899 frogsaccording to their fitness order The frogs are sortedin descending order according to their fitness Thefirst frog is assigned to the first memeplex the secondfrog is assigned to the second memeplex the 119898thfrog is assigned to the 119898th memeplex the (119898 +1)th frog is assigned back to the first memeplexand so on The best and worst positions of eachfrog in each memeplex are indicated as 119865

119887and 119865

119908

respectively and the frog in the best position in theentire population is indicated as 119865

119892

(3) Memetic evolution during the memetic evolution ofthe memeplex the frog in the worst position 119865

119908is

updated through a local search in its memeplex Thenew position of the worst frog is updated accordingto (1) to (3)

119863 = 119872(rand (0 1) lowast (119865119887 minus 119865119908)) (1)

119865119908new = 119865119908old + 119863 (minus119863max le 119863 le 119863max) (2)

119863 = 119872(rand (0 1) lowast (119865119892 minus 119865119908)) (3)

where119863 is themoving distancematrix119872(rand(0 1))is the random number matrix within the range [0 1]and 119863max is the maximum distance that the frog ispermitted to moveThe memetic evolution procedure in SFLA is illus-trated in Figure 2If the new position of the worst frog is better than itsprevious position after (1) and (2) are calculated thenit replaces the position of the worst frog and the worstposition 119865

119908is recalculated Otherwise substitute 119865

119892

for 119865119887and repeat the worst position of the frog

updating calculations in (3) and (2) If the frog in theworst position still cannot obtain a better positionthen a frog with a new position is stochasticallyproduced to replace the frog in the worst position 119865

119908

Consequently each memeplex follows a predefinedmemetic evolutionary time

(4) Global shuffling process all frogs are mixed andsorted in descending order according to their fitnessThe memeplex is then divided according to the neworder and then Step (2) is repeated

(5) Local evolution and global shuffling processes con-tinue until convergence is achieved

4 ASP Problem Statement

The final goal of ASP is to enhance assembly efficiency andreduce assembly difficulties and costs Several parametersare involved to achieve this objective This study employsseveral essential parameters as evaluating indicators includ-ing geometric feasibility assembly stability changing times

Start

Yes

End

No

More than the predefined memetic evolutionary times

No

Yes

No

Yes

Find Fw Fb Fg

Update Fw by Eqs (1) (2)

New Fw is better than old Fw

New Fw is better than old Fw

Stochastically produces a new Fw

New Fw replaces Fb

Fg replaces update Fw by (3) and (2)Fb

Figure 2 Memetic evolution procedure in SFLA

of assembly tool and changing times of assembly directionThe fitness function is ultimately developed through theevaluating indicators

41 Geometric Feasibility The assembly direction is dividedinto six types of direction as follows 119889(119896) = +119909 +119910 +119911 minus119909minus119910 minus119911 Interference values 119868

119894119895119889119896(119896 = 1 2 6 and 119889

119896isin

119889(119896)) describe whether part 119875119894interfere with part 119875

119895when

moving along 119889119896direction 119868

119894119895119889119896is as follows

119868119894119895119889119896

=

0 119901119894does not interfere with 119901

119895in the 119889

119896direction

1 119901119894interferes with 119901

119895in the 119889

119896direction

(4)

Suppose that AP = (1198751 1198752 119875

119899) is an assembly

sequence The part set AP1= 1198751 1198752 119875

119894minus1 is the set in

which parts have been assembled and 119875119894is the part to be

assembled Then 119878119896(119875119894) (119896 = 1 2 6) is the sum of the

Science and Technology of Nuclear Installations 5

IF 119888119875119894119875119895= 2 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSEIF 119888119875119894119875119895== 0 119895 isin [1 119894 minus 1] THEN

Assembly operation is unstableELSE

IF 119904119875119895119875119894= 1 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSE IF 119904

119875119895119875119894== 0 THEN

Assembly operation is unstableENDIF

ENDIF

Algorithm 1 Stability evaluation method of the assemblysequence

interference values which is 119875119894and each part in AP

1when 119875

119894

is assembled along 119889119896direction 119878

119896(119875119894) is as follows

119878119896(119875119894) =

119894minus1

sum

119895=1

119868119875119894119875119895119889119896 (5)

If 119878119896(119875119894) = 0 then 119875

119894can be assembled along 119889

119896

Otherwise 119875119894cannot be assembled along 119889

119896 In this case we

obtain119863119862(119875119894) = 119889

119896| 119878119896(119875119894) = 0 that is the feasible assembly

direction set of 119875119894 For each 119875

119894(1 lt 119894 le 119899) if 119863119862(119875

119894) = 120601

then AP is the feasible assembly sequence otherwise AP isthe infeasible assembly sequence 119899

119891is expressed as the total

times of assembly interference of APThe value of 119899119891is equal

to the total times of119863119862(119875119894) = 120601 (1 lt 119894 le 119899) in AP

42 Assembly Stabilities In the actual assembly process partsmay become unstable because of gravity Several assemblyoperationsmust use a jig or auxiliary tool tomaintain stabilitywhen a part is unstable during the assembly process whichresults in an inefficient assembly Therefore the stability ofthe assembly sequence should be evaluated

The augmented adjacencymatrix119862 = (119888119894119895)119899times119899

and supportmatrix 119878 = (119904

119894119895)119899times119899

are defined to evaluate the stability of theassembly sequence In the augmented adjacency matrix 119888

119894119895

expresses the connection type between 119875119894and 119875119895 For a stable

connection 119888119894119895= 2 for a contact connection 119888

119894119895= 1 and

for a noncontact connection 119888119894119895= 0 In the support matrix

119904119894119895expresses the support type between 119875

119894and 119875

119895 For stable

support 119904119894119895= 1 otherwise 119904

119894119895= 0

Suppose that AP = (1199011 1199012 119901

119899) is an assembly

sequence The part set AP1= 1199011 1199012 119901

119894minus1 is expressed

as the parts having been assembled and 119875119894is expressed as the

part to be assembled The stability evaluation method of theassembly sequence is shown in Algorithm 1 In this study 119899

119904

expresses the times of the assembly sequence stable operationA smaller 119899

119904indicates a more stable assembly sequence

43 Changing Times of Assembly Tool Given the particularityof each assembly part different assembly tools should be usedin the actual assembly process Changing the assembly tool

leads to a long assembly time and high cost for the assemblyprocessTherefore changing times of assembly tool should beas few as possible

Suppose the assembly sequence is AP = (1198751 1198752 119875

119899)

and assembly tool sequence of AP is 119879119888 119879119888(119875119894) is expressed

as the assembly tool of 119875119894 Assembly tool of each part is

determined by the characteristic of each part and the availableassembly tool The assembly tool sequence for an AP as wellas the optimal assembly tool sequence is predeterminedChanging times of assembly tool 119899

119905are calculated as shown

in Algorithm 2

44 Changing Times of Assembly Direction The reducedchanging times of assembly direction shorten assembly timeand enhance assembly efficiency Supposing the assemblysequence is AP = (119875

1 1198752 119875

119899) changing times of assembly

direction 119899119889are calculated as shown in Algorithm 3

45 Fitness Function Different radioactive equipment undervarious environmentsmay have varying influence degrees forthe evaluating indicators Therefore weighting factors mustbe determined according to the actual situation A penaltyfunction 119888

119891119899119891is applied to infeasible assembly sequence to

speed up the algorithm convergence rate Then the weightedfitness function is as follows

119891 = 119888119904119899119904+ 119888119905119899119905+ 119888119889119899119889+ 119888119891119899119891 (6)

where 119888119904 119888119905 119888119889 and 119888

119891are the weighting factors for each

evaluating indicator and 119888119891must be generally larger than the

other three weighting factors (ie 119888119891ge (1198992)max119888

119904 119888119905 119888119889)

In this study a small fitness function value indicates goodposition of the frog and good assembly sequence

5 Improved SFLA for ASP

51 Local Search Strategy Based on a Swap Sequence ASP is acombinatorial optimization problem in which each solutiondimension is discrete A GA can solve the discrete optimiza-tion problem by using crossover andmutation operatorsTheimproved SFLA introduces a local search strategy based on aswap sequence to address this problem

511 Swap Factor and Swap Sequence

(1) Swap Factor Suppose that an assembly sequence thatincludes 119899 parts is expressed as AP = (119875

1 1198752 119875

119899) The

function of swap factor V119900(120574 120596) is to swap the positions of120574 and 120596 to form a new assembly sequence For example if theinitial assembly sequence is AP = (2 4 3 5 1) and the swapfactor is V119900 = V119900(2 4) then AP1015840 = AP oplus V119900 = (2 5 3 4 1)oplus indicates that the swap factor is acting on the assemblysequence

(2) Swap Sequence V119900119904 = (V1199001 V1199002 V119900

119899) expresses a

swap sequence that consists of 119899 swap factors in whichV1199001 V1199002 V119900

119899are the swap factors and their order does not

satisfy the commutative lawThe effect of a swap sequence on

6 Science and Technology of Nuclear Installations

Step 1 Set 119894 = 0119898 = 1 119899119905= 0

Step 2 IF ⋂119894+119898119894119879119888(119875119894) = THEN

lowastAssembly tool must be changed when 119875119894is to be assembledlowast

119899119905= 119899119905+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899 ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119905

Algorithm 2 Calculation procedure for changing times of assembly tool

Step 1 Set 119894 = 0119898 = 1 119899119889= 0

Step 2 IF ⋂119894+119898119894119863119888(119875119894) = THEN

lowastAssembly direction must be changed when 119875119894is to be assembledlowast

119899119889= 119899119889+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119889

Algorithm 3 Calculation procedure for changing times of assembly direction

an assembly sequence is equal to the effect of each swap factorin a swap sequence on the assembly sequence

AP119886andAP

119887are two assembly sequences V119900119904(AP

119887ΘAP119886)

expresses the swap sequence in which AP119886is adjusted as AP

119887

It can be expressed as (5)AP119887= AP119886oplus V119900119904 (AP

119887ΘAP119886) = AP

119886oplus (V1199001 V1199002 V119900

119899)

(7)A false code of the swap sequence is shown in

Algorithm 4 For example if AP119886= (1 4 2 5 3) and AP

119887=

(2 3 5 1 4) then the swap sequence is V119900119904(AP119887ΘAP119886) =

(V1199001(1 3) V119900

2(2 5) V119900

3(3 4))

512 Frog Position Updating Strategy 119863 is the number ofswap factors contained by the moving distance matrix Dceil is the top integral function 119903119863 (119903 isin [0 1]) is the firstceil(119903119863) swap factors in the swap sequenceD For exampleif 119863 = V119900119904 = (V119900

1 V1199002 V1199003 V1199004) and 119903 = 06 then 119863 = 4

and 119903119863 = (V1199001 V1199002 V1199003)

Science and Technology of Nuclear Installations 7

Step 1 Set 119894 = 1 119895 = 0Step 2 IF AP

119886(119894) = AP

119887(119894) THEN

119895 = 119895 + 1

V = 119891119894119899119889(AP119887== AP

119886(119894))

V119900119895= V119900(119894 V)

AP119886= AP119886oplus V119900119895

ENDIFStep 3 Set 119894 = 119894 + 1

IF 119894 le 119899 minus 1 THENProceed to Step 2

ELSEProceed to Step 4

ENDIFStep 4 V119900119904 = (V119900

1 V1199002 V119900

119895)

Algorithm 4 False code of the swap sequence

Equations (1) to (3) are improved as follows

119863 = 119872(rand (0 1) lowast (119865119887Θ119865119908)) (8)

119865119908new = 119865119908old oplus 119863 (

1003817100381710038171003817119863min1003817100381710038171003817 le 119863 le

1003817100381710038171003817119863max1003817100381710038171003817) (9)

119863 = 119872(rand (0 1) lowast (119865119892Θ119865119908)) (10)

52 Diversity Control Strategy After population sorting inSFLA the grouping operator makes the best frog positionsimilar in each memeplex when the first 119898 frogs satisfy 119865

1=

1198652= sdot sdot sdot = 119865

119898(119898 is the quantity of the memeplex) Based on

(8) to (10) eachmemeplex can easily converge to the best frogposition 119865

119892of the entire population The algorithm search

space and the probability of algorithm convergence with theglobally optimal solution are reduced This study proposesa diversity control strategy to avoid homoplasy The controlpolicy is as follows

(1) Compute119873 of the preceding same frog position afterthe grouping operator

(2) If119873 lt 119898 then proceed to step (4)(3) The next population is based on standard GA(4) The population is based on the other SFLA steps

53 Improved SFLA Steps The basic steps to solve the ASPproblem by using the improved SFLA are shown in Figure 3The detailed steps are as follows

(1) Parameter initialization and population initializationfrog population size is size The number of frogmemeplex is 119898 The population iterative is iter Thelocal search iteration is mrun The maximum andminimum frog moving distances are 119863max and119863min respectively The crossover probability is pcThe adaptive mutation probability is pm

(2) Modified grouping operator the frogs are sorted indescending order according to their fitness during thepreprocessing of the grouping operator in SFLA In

Start

(3) Memetic evolution

(5) Global shuffling process

Yes

(2) Modified grouping operator

(1) Population initialization and parameter initialization

Convergence criteria are satisfied

Output the optimal solution

End

No

(4) Diversity control strategy

(local search based on swap sequence)

Figure 3 Memetic evolution procedure in SFLA

ASP a small fitness function value indicates good frogposition and good assembly sequence Therefore thegrouping operator should be modified The modifiedgrouping operator is as follows Suppose that the scaleof the frog population is 119873 which is then dividedinto 119898 memeplexes All frogs in the population arearranged in ascending order according to the fitnessfunction valueThefirst frog enters the firstmemeplexand the second frog enters the secondmemeplex andso on until the 119898 frog enters the 119898th memeplexThe (119898 + 1)th frog is then assigned back to thefirst memeplex and so on All individual frogs areassigned according to the aforementioned rule

(3) Memetic evolution (local search based on a swapsequence) in the improved SFLA memetic evolutionis modified and performed by using a local searchstrategy based on a swap sequence until the mrungeneration

(4) Optimal sampling different strategy the optimal sam-pling different strategy is included in the improvedSFLA to avoid homoplasy for each memeplex

(5) Global shuffling process the global shuffling processof the improved SFLA is similar to SFLA and updatesthe best position 119865

119892of the frog population

(6) The next step is determining whether the iterationshould be terminated according to the terminal con-dition of the algorithm If the terminal condition is

8 Science and Technology of Nuclear Installations

Table 1 Part assembly tool sets

Part number Part name Assembly tool1 Nut washer assembly 1 1198791

2 Nut washer assembly 2 1198791

3 Nut washer assembly 3 1198791

4 Nut washer assembly 4 1198791

5 Hydraulic cylinder 1198794

6 Pole 1 1198793

7 Pole 2 1198793

8 Pole 3 1198793

9 Pole 4 1198793

10 Strut 1 1198794

11 Nut 1 1198791

12 Nut 2 1198791

13 Nut 3 11987911198792

14 Bolt 1 11987911198793

15 Bolt 2 11987911198793

16 Pin 1 1198793

17 Nut 4 11987911198792

18 Pin 2 1198793

19 Central pin 1198793

20 Back plate 1198794

21 Strut 2 1198794

22 Nut washer assembly 5 1198791

23 Nut washer assembly 6 1198791

24 Nut washer assembly 7 1198791

25 Nut washer assembly 8 1198791

26 Axis 1198793

27 Hydraulic pressure scissors 1198794

28 Hydraulic pressure shear blades 1 1198795

29 Hydraulic pressure shear blades 2 1198795

satisfied then the iteration ends Otherwise return tostep (2)

6 Experiment and Analysis

The application program based on the improved SFLAis compiled under MATLAB environment The computerenvironment of the application program consists of a 20GHzCPU 2GBmemory andWindows 7 32-bit operating systemThe hydraulic pressure shear which contains 29 parts is usedfor the ASP experiment The exploded view of the hydraulicpressure shear is shown in Figure 4 The components of theassembly tool sets are listed in Table 1

61 ASP Experiment Based on SFLA After conducting anorthogonal experiment on the assembly of the hydraulicpressure shear the algorithm rapidly identifies an optimalassembly sequence when the weighting factors of the eval-uating indicator in the fitness function are 119888

119891= 4 119888

119904=

05 119888119905= 02 and 119888

119889= 03 If the memeplex has few local

search iterations then it also undergoes few evolution timeswhich reduces information exchange within the memeplex

If the memeplex has many local search iterations then itundergoes multiple local searches which increases algorithmsearch time and makes the best frog position of variousmemeplexes similar It also causes the algorithm to carry outGA several times and thus slows down the convergence rate ofthe algorithm If the maximum moving distance 119863max is toosmall then the global algorithm search capability is reducedIt causes the algorithm to easily fall into a local search If119863maxis too large then the algorithm is unable to find the globallyoptimal solutionAftermultiple comparison experiments thealgorithm optimization capability is observed to be optimalwhen the local search iteration of the memeplex is 10crossover probability is 08 adaptive mutation probabilityis 01 maximum moving distance is 8 minimum movingdistance is 1 and frog quantity in thememeplex ismaintainedat 30

The ASP experiment is conducted with population sizes60 120 180 and 240 given that each parameter value of theimproved SFLA and the weighting factor of the evaluatingindicator in the fitness function are the same The fitnessfunction value distributed the optimal assembly sequencefrom the results of the ASP experiment The analysis resultsare shown in Figure 5 in which the number of algorithmiterations is 600 and that of repeating operation times is 50A lot of experiments show that the fitness value of the globaloptimal assembly sequence 119865 is 21 As shown in Figure 5the distributed situation of the local optimal fitness value iswithin the following ranges 21 to 30 31 to 40 41 to 5051 to 60 and gt60 As shown in Figure 5 the distributedsituations of fitness value of local optimal assembly sequencediffer along with various population sizes When populationsize is 60 and time of experiment is 50 there is only onefitness value of local optimal assembly sequence in sections21 to 30 As population size increases the quantity offitness value of local optimal assembly sequence identifiedby the algorithm in this section gradually increases Whenpopulation size increases to 240 the quantity identifiedby the algorithm in this section is 20 As population sizeincrease the quantity of outstanding assembly sequenceswhose fitness value is smaller gradually increases As shownin Table 2 the increase in algorithm population size reducesalgorithm iteration efficiency and the operation time of thealgorithm is extended In this experiment the probabilityof the global optimal assembly sequence identified by thealgorithm is highest when population size is 240 and theaverage consuming time of a single experiment is in the rangeof acceptable with 515S

The mean and optimal average fitness of iteration in50 experiments when population size is 120 are shown inFigure 6 As algorithm iteration increases the optimal fitnessaverage value fluctuates However the optimal and averagevalues of the average fitness steadily decrease from the overalltendency which shows that the stability of the algorithm isgood and the algorithm convergence rate at the later periodis slow

The mean and optimal average fitness of 1 of the 50experiments in which the global optimal assembly sequenceis obtained when population size is 120 is shown in Figure 7In the algorithm implementation the mean fitness exhibits a

Science and Technology of Nuclear Installations 9

1

2

3 45

6

7

89

10

11

12

13

14

1516

1718

19

20

21

22

23

24

25

26

2728

29

Z

X

Y

Figure 4 Exploded view of the hydraulic pressure shear

huge fluctuation because the optimal sample is homoplasyAccording to the diversity control strategy the algorithmexecutes GA and increases average fitness The homoplasy ofthe algorithm improves after executing GA

62 Algorithm Comparison Experiment An algorithm com-parison experiment is conducted among improved SFLA(SFLA-GA) GA SFLA PSO and AMPSO [7] to verify theperformance of the improved SFLA for the ASP problemAMPSO ismodifiedmethod for ASP in RHM in our previouswork [7]

A hydraulic pressure shear is employed to carry out ASPwith the same programming and PC environments as thoseindicated in Section 61 Moreover the parameter 119891

119889= 21

and values of other related parameters in AMPSO are thesame as those presented in [11] Inertia weight in PSO is 06and the values of the other parameters are similar to thoseof AMPSO The experiment results with population sizes of60 and 240 are shown in Table 3 The algorithm convergencecurves of different population sizes are shown in Figures 8and 9 (ie variation of population average fitness values alongwith the iterations)

As shown in Table 3 the probability of a feasible assemblysequence identified by the improved SFLA is enhanced whenpopulation size increasesThe probability is higher than thosefor SFLA and GA but lower than those for AMPSO andPSO Under the same population size the improved SFLAexhibits a superior assembly sequence than those of GA andSFLA The value of the optimal assembly sequence fitnessfunction identified by the improved SFLA is less than those

identified by GA and SFLA When population size is 60the improved SFLA obtains an acceptable assembly sequenceand the fitness function value is 29 However GA and SFLAare unable to obtain an acceptable assembly sequence evenwhen the population size is 240 The fitness values of theoptimal assembly sequence are similar in PSO and AMPSOThe execution time of the improved SFLA is slightly less thanthat of GA and significantly less than those of AMPSO andPSO Consequently the efficiency of the proposed algorithmis acceptable Based on the local optimal fitness averagevalue of the algorithm the improved SFLA exhibits a higherconvergence rate than those of GA and SFLA and is nearthose of AMPSO and PSO The improved SFLA is evenbetter than PSO when population size is 240 Therefore theoptimization capability efficiency and convergence rate ofthe improved SFLA are better than those of GA whereasits optimization capability and convergence rate are betterthan those of SFLA The overall performance of the SFLA-GA for solving ASP problems proposed in this study issimilar to those of AMPSO and PSO As shown in Figures8 and 9 the stochastic initializing population qualities ofthe five algorithms are approximately similar Therefore thepreceding analysis is reliable

7 Conclusions

RHM is an important mean of ensuring the reliabilityof radioactive equipment and has a wide application inradioactive installations RHMP predetermine the mainte-nance procedures of radioactive equipment during the design

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

2 Science and Technology of Nuclear Installations

for assembly Finding an optimal assembly sequence to satisfytime cost and reliability requirements is a combinatorialproblem the complexity of this problem is proportional tothe number of equipment parts The number of feasibleRHM sequences increases with equipment complexity [9] Alarge number of equipment parts results in a combinationexplosion and the optimal solution is omitted ASP is thenshown to be NP-complete [10]

Optimization techniques based on principles inspired bynatural systems have been proposed over the past decades tosolve the combinatorial explosion problem [11] The shuffledfrog leaping algorithm (SFLA) is a recent metaheuristicoptimization algorithm that is inspired by the memeticevolution of a group of frogs when seeking food SFLAinvolves a set of frogs that cooperate to achieve a unifiedbehavior for the entire system which produces a robustsystem that can find high-quality solutions to problemswith large search spaces [12] SFLA exhibits global searchingcapability rapid convergence and strong robustness It hasbeen successfully applied to several fields However SFLAis suitable for continuous optimization [13] SFLA should beimproved when applied to ASP which is a discrete search andoptimization problem

In this study an improved SFLA is presented to solvethe ASP problem for RHM in radioactive environment EachSFLA operation is redefined In particular a swap operatorand a swap sequence are introduced and the local searchstrategy is designed to directly search the discrete domainA diversity control strategy based on genetic algorithm (GA)is proposed to improve the search for global optimal solu-tions An ASP experiment show that the algorithm exhibitsoutstanding performance in solving the ASP problem

The remainder of the paper is organized as followsSection 2 introduces related works Section 3 describes SFLASection 4 states the ASP problem Section 5 discusses theimproved SFLA for ASP Section 6 describes the experimentsand the analyses Finally Section 7 provides the conclusionsof the study

2 Related Studies

21 RHM Maintenance preserves or restores a system orfacility to its desired state The following problems shouldbe considered for maintenance in radioactive installation (1)safety of the maintenance worker in cases where humanscannot gain access because of the high radiation dose rate (2)feasibility of maintenance work in cases where humans hasdifficulty working with equipment because of certain condi-tions (eg small spaces and narrow gaps) and (3) reliabilityofmaintenance work in cases where harsh environments andheavy workloads cause human errors

RHM is applied to solve the aforementioned problems[3ndash5] RHM mainly repairs fault parts that cause equipmentto stop working in a radioactive environment Operationsof RHN mainly include replacement and disposal workswhich are remotely handled by using power and master-slavemanipulators [14]The following observations aremade(1) RHM differs from conventional equipment maintenance

because it employs a robot or a remote operation tool in a hotcell instead of a human (2) The robot must be teleoperatedfully controlled or supervised by a human outside the hot cellbecause the majority of RHM tasks require the intuition andintelligence of a human [5] (3) A human does not need tobe physically present at the work site to conduct maintenancework

RHM will be applied to radioactive installation suchas the China Fusion Engineering Test Reactor [15] theInternational Thermonuclear Experimental Reactor (ITER)[6] and the European Organization for Nuclear Research[16]

22 Robots in RHM Several robots have been developed forRHM Takeda et al [17] designed three kinds of robots thatcan transport different parts in a radioactive environmentThe French Atomic Energy Agency Interactive RoboticsLaboratory developed an industrial robot system for a nuclearspent fuel reprocessing plant [18] The robot which usesRX170 as a slave arm and a control platform called TAO2000V2 supports a master-slave operation with a force feedbackand tolerates radiation up to a 10 kGy integrated dose [19]Sanders [20] developed a remote handling system with ldquomanin the looprdquo approach that provides the remote robot operatorfor the Joint European Torus Vale et al [21] developedan autonomous mobile robot for ITER Terada et al [22]designed and developed a pick-and-place work robot tocope with the module placement for the semiconductortracker barrel assembly The robot can place modules with amechanical precision of over 25 120583m Lee et al [23] developeda bridge transported servo-manipulator system to overcomethe limited workspace of conventional mechanical master-slave manipulators in a hot cell Lee et al [24] developeda cable-driven dual arm master-slave servo-manipulator forthe pyroprocess research facility

23 RHMP In radioactive installation RHMP predeter-mines the maintenance process during the stages of radioac-tive installation design Traditional RHM is mainly per-formed through an empirical design and physical verificationby experiment platform It is unsuitable for complex structureinstallation and can be laborious and ineffective The experi-ment platform results in high costs and a long cycle

The development of computer artificial intelligence andsimulation technologies allows the application of virtualmaintenance planning to RHM [25] Takeda et al [26]developed a virtual reality simulator to support the Banketsimulation of theRHMprocess of ITERHeemskerk et al [27]studied the simulation process dynamics based on the ITERRHM simulator Geng et al [28] developed a novel virtualmaintenance application for maintenance safety evaluationto provide recommendations on maintenance safety Esqueet al [29] completed a digital simulation model of the ITERseparator Shuff et al [30] developed a set of discrete eventsimulation tools for the remote operating process planning ofthe ITER hot cabinet Robbins et al [31] achieved a real-timeand visualized track of remote operating process planning byusing virtual reality and an intelligent database Park et al [32

Science and Technology of Nuclear Installations 3

33] studied visualization and simulation of a nuclear facilitydisassembling process and established an RHM system

In the aforementioned studies maintenance sequencewas generally obtained in an exploratory manner because ofthe lack of a guided optimized maintenance sequence Suchsequencemay be inefficient and the optimal solutionmight beomitted without intelligent support Introducing intelligentplanning technologies to ASP is essential to further enhancevirtual maintenance planning [7]

24 Intelligent ASP for Complex Products Assembling prod-ucts (by humans or by robots) is the act of combiningparts of equipment ASP obtains the order for each part andsubassembly which is then inserted into an incrementallyexpanding subassembly that eventually leads to a final assem-bly ASP is a combinatorial problem in complex productsSolving this problem with human involvement is difficultand impractical because of the combinatorial explosion issueResearch in intelligence assembly sequencing has rapidlyincreased in recent years The intelligence ASP problem isregarded as a discrete search and optimization problemVarious artificial intelligence approaches have been proposedrecently including graph theory [34 35] subassembly detec-tion [36 37] motion planning [38] and evolution algorithm[11]

Evolution algorithms provide new solutions to vari-ous complex optimization problems by imitating the self-organization mechanism of natural biological communitiesand the adaptive ability of evolution Chen and Xiao [39]developed an enhancing artificial bee colony algorithmwith self-adaptive searching strategy and artificial immunenetwork operators for global optimization Xu et al [40]developed an improved genetic algorithm for distributionnetwork planning Chen and Ju [41] developed a novelartificial bee colony algorithm for solving the supply chainnetwork design under disruption scenarios Cheng et al[42] developed a metaheuristics for airport gate assignmentLorpunmance and Sap [43] developed an ant colony opti-mization for dynamic job scheduling in grid environmentEvolution algorithms such as GA [44ndash46] the ant colonyalgorithm [47 48] the particle swarm algorithm [49ndash52] andthe artificial bee colony algorithm [53] have been studiedrecently in ASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems and hasbeen found to be effective in searching for global solutionsRahimi-Vahed andMirzaei proposed a hybrid multiobjectiveSFLA for a mixed model assembly line sequencing prob-lem [54] and bicriteria permutation flow shop schedulingproblem [55] Li et al [56] proposed an effective SFLA fora multiobjective flexible job shop scheduling problems Fangand Wang [57] introduced an effective SFLA for a resource-constrained project scheduling problem Li et al [58] devel-oped a modified SFLA for continuous optimization Theaforementioned studies show that SFLA is a simple robustand fast algorithm for solving combinatorial optimizationproblems

Start

(3) Memetic evolution (by local search)

(4) Global shuffling process

Yes

(2) Grouping operation

(1) Population initialization and parameter initialization

Convergence criterions are satisfied

Output the optimal solution

End

No

Figure 1 SFLA procedure

3 SFLA

SFLA is a metaheuristic optimization method that identifiessolutions by simulating collaboration behaviors and inter-active information similar to a frog community searchingfor food in a natural environment This algorithm dividespopulation into several subpopulations and the evolutionof memes is driven by the exchange of global informationamong subpopulations and the local evolutionary searchwithin subpopulation

SFLA is described in detail as follows The populationconsists of several frogs and each frog is a solution to theproblem The population is divided into several subpopula-tions through a grouping operator to simulate frog groupingbehaviors Each subpopulation is called a memeplex Amemeplex is composed of frogs with the same meme thatperform local searches Each frog has its own idea but isalso influenced by other frogs in the same memeplex Frogsadjust their positions through memetic evolution After apredefined number of memetic evolutionary times frogs indifferent subpopulations exchange information through aglobal shuffling process The alternating memetic evolutionand global shuffling process make the frogs jump out of thelocal optimum and evolve toward the global optimum

The SFLAprocedure is illustrated in Figure 1Thedetailedprocedure is as follows

(1) Population initialization and parameter initializationSFLA initially creates a population of 119865 frogs as acertain solution amount The 119894th frog (ie the 119894thsolution to the problem) is represented as 119865

119894=

4 Science and Technology of Nuclear Installations

(1198911198941 1198911198942 119891

119894119904) where 119904 is the solution space dimen-

sion(2) Grouping operator a grouping operator separates119865 frogs into 119898 memeplex which contains 119899 frogsaccording to their fitness order The frogs are sortedin descending order according to their fitness Thefirst frog is assigned to the first memeplex the secondfrog is assigned to the second memeplex the 119898thfrog is assigned to the 119898th memeplex the (119898 +1)th frog is assigned back to the first memeplexand so on The best and worst positions of eachfrog in each memeplex are indicated as 119865

119887and 119865

119908

respectively and the frog in the best position in theentire population is indicated as 119865

119892

(3) Memetic evolution during the memetic evolution ofthe memeplex the frog in the worst position 119865

119908is

updated through a local search in its memeplex Thenew position of the worst frog is updated accordingto (1) to (3)

119863 = 119872(rand (0 1) lowast (119865119887 minus 119865119908)) (1)

119865119908new = 119865119908old + 119863 (minus119863max le 119863 le 119863max) (2)

119863 = 119872(rand (0 1) lowast (119865119892 minus 119865119908)) (3)

where119863 is themoving distancematrix119872(rand(0 1))is the random number matrix within the range [0 1]and 119863max is the maximum distance that the frog ispermitted to moveThe memetic evolution procedure in SFLA is illus-trated in Figure 2If the new position of the worst frog is better than itsprevious position after (1) and (2) are calculated thenit replaces the position of the worst frog and the worstposition 119865

119908is recalculated Otherwise substitute 119865

119892

for 119865119887and repeat the worst position of the frog

updating calculations in (3) and (2) If the frog in theworst position still cannot obtain a better positionthen a frog with a new position is stochasticallyproduced to replace the frog in the worst position 119865

119908

Consequently each memeplex follows a predefinedmemetic evolutionary time

(4) Global shuffling process all frogs are mixed andsorted in descending order according to their fitnessThe memeplex is then divided according to the neworder and then Step (2) is repeated

(5) Local evolution and global shuffling processes con-tinue until convergence is achieved

4 ASP Problem Statement

The final goal of ASP is to enhance assembly efficiency andreduce assembly difficulties and costs Several parametersare involved to achieve this objective This study employsseveral essential parameters as evaluating indicators includ-ing geometric feasibility assembly stability changing times

Start

Yes

End

No

More than the predefined memetic evolutionary times

No

Yes

No

Yes

Find Fw Fb Fg

Update Fw by Eqs (1) (2)

New Fw is better than old Fw

New Fw is better than old Fw

Stochastically produces a new Fw

New Fw replaces Fb

Fg replaces update Fw by (3) and (2)Fb

Figure 2 Memetic evolution procedure in SFLA

of assembly tool and changing times of assembly directionThe fitness function is ultimately developed through theevaluating indicators

41 Geometric Feasibility The assembly direction is dividedinto six types of direction as follows 119889(119896) = +119909 +119910 +119911 minus119909minus119910 minus119911 Interference values 119868

119894119895119889119896(119896 = 1 2 6 and 119889

119896isin

119889(119896)) describe whether part 119875119894interfere with part 119875

119895when

moving along 119889119896direction 119868

119894119895119889119896is as follows

119868119894119895119889119896

=

0 119901119894does not interfere with 119901

119895in the 119889

119896direction

1 119901119894interferes with 119901

119895in the 119889

119896direction

(4)

Suppose that AP = (1198751 1198752 119875

119899) is an assembly

sequence The part set AP1= 1198751 1198752 119875

119894minus1 is the set in

which parts have been assembled and 119875119894is the part to be

assembled Then 119878119896(119875119894) (119896 = 1 2 6) is the sum of the

Science and Technology of Nuclear Installations 5

IF 119888119875119894119875119895= 2 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSEIF 119888119875119894119875119895== 0 119895 isin [1 119894 minus 1] THEN

Assembly operation is unstableELSE

IF 119904119875119895119875119894= 1 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSE IF 119904

119875119895119875119894== 0 THEN

Assembly operation is unstableENDIF

ENDIF

Algorithm 1 Stability evaluation method of the assemblysequence

interference values which is 119875119894and each part in AP

1when 119875

119894

is assembled along 119889119896direction 119878

119896(119875119894) is as follows

119878119896(119875119894) =

119894minus1

sum

119895=1

119868119875119894119875119895119889119896 (5)

If 119878119896(119875119894) = 0 then 119875

119894can be assembled along 119889

119896

Otherwise 119875119894cannot be assembled along 119889

119896 In this case we

obtain119863119862(119875119894) = 119889

119896| 119878119896(119875119894) = 0 that is the feasible assembly

direction set of 119875119894 For each 119875

119894(1 lt 119894 le 119899) if 119863119862(119875

119894) = 120601

then AP is the feasible assembly sequence otherwise AP isthe infeasible assembly sequence 119899

119891is expressed as the total

times of assembly interference of APThe value of 119899119891is equal

to the total times of119863119862(119875119894) = 120601 (1 lt 119894 le 119899) in AP

42 Assembly Stabilities In the actual assembly process partsmay become unstable because of gravity Several assemblyoperationsmust use a jig or auxiliary tool tomaintain stabilitywhen a part is unstable during the assembly process whichresults in an inefficient assembly Therefore the stability ofthe assembly sequence should be evaluated

The augmented adjacencymatrix119862 = (119888119894119895)119899times119899

and supportmatrix 119878 = (119904

119894119895)119899times119899

are defined to evaluate the stability of theassembly sequence In the augmented adjacency matrix 119888

119894119895

expresses the connection type between 119875119894and 119875119895 For a stable

connection 119888119894119895= 2 for a contact connection 119888

119894119895= 1 and

for a noncontact connection 119888119894119895= 0 In the support matrix

119904119894119895expresses the support type between 119875

119894and 119875

119895 For stable

support 119904119894119895= 1 otherwise 119904

119894119895= 0

Suppose that AP = (1199011 1199012 119901

119899) is an assembly

sequence The part set AP1= 1199011 1199012 119901

119894minus1 is expressed

as the parts having been assembled and 119875119894is expressed as the

part to be assembled The stability evaluation method of theassembly sequence is shown in Algorithm 1 In this study 119899

119904

expresses the times of the assembly sequence stable operationA smaller 119899

119904indicates a more stable assembly sequence

43 Changing Times of Assembly Tool Given the particularityof each assembly part different assembly tools should be usedin the actual assembly process Changing the assembly tool

leads to a long assembly time and high cost for the assemblyprocessTherefore changing times of assembly tool should beas few as possible

Suppose the assembly sequence is AP = (1198751 1198752 119875

119899)

and assembly tool sequence of AP is 119879119888 119879119888(119875119894) is expressed

as the assembly tool of 119875119894 Assembly tool of each part is

determined by the characteristic of each part and the availableassembly tool The assembly tool sequence for an AP as wellas the optimal assembly tool sequence is predeterminedChanging times of assembly tool 119899

119905are calculated as shown

in Algorithm 2

44 Changing Times of Assembly Direction The reducedchanging times of assembly direction shorten assembly timeand enhance assembly efficiency Supposing the assemblysequence is AP = (119875

1 1198752 119875

119899) changing times of assembly

direction 119899119889are calculated as shown in Algorithm 3

45 Fitness Function Different radioactive equipment undervarious environmentsmay have varying influence degrees forthe evaluating indicators Therefore weighting factors mustbe determined according to the actual situation A penaltyfunction 119888

119891119899119891is applied to infeasible assembly sequence to

speed up the algorithm convergence rate Then the weightedfitness function is as follows

119891 = 119888119904119899119904+ 119888119905119899119905+ 119888119889119899119889+ 119888119891119899119891 (6)

where 119888119904 119888119905 119888119889 and 119888

119891are the weighting factors for each

evaluating indicator and 119888119891must be generally larger than the

other three weighting factors (ie 119888119891ge (1198992)max119888

119904 119888119905 119888119889)

In this study a small fitness function value indicates goodposition of the frog and good assembly sequence

5 Improved SFLA for ASP

51 Local Search Strategy Based on a Swap Sequence ASP is acombinatorial optimization problem in which each solutiondimension is discrete A GA can solve the discrete optimiza-tion problem by using crossover andmutation operatorsTheimproved SFLA introduces a local search strategy based on aswap sequence to address this problem

511 Swap Factor and Swap Sequence

(1) Swap Factor Suppose that an assembly sequence thatincludes 119899 parts is expressed as AP = (119875

1 1198752 119875

119899) The

function of swap factor V119900(120574 120596) is to swap the positions of120574 and 120596 to form a new assembly sequence For example if theinitial assembly sequence is AP = (2 4 3 5 1) and the swapfactor is V119900 = V119900(2 4) then AP1015840 = AP oplus V119900 = (2 5 3 4 1)oplus indicates that the swap factor is acting on the assemblysequence

(2) Swap Sequence V119900119904 = (V1199001 V1199002 V119900

119899) expresses a

swap sequence that consists of 119899 swap factors in whichV1199001 V1199002 V119900

119899are the swap factors and their order does not

satisfy the commutative lawThe effect of a swap sequence on

6 Science and Technology of Nuclear Installations

Step 1 Set 119894 = 0119898 = 1 119899119905= 0

Step 2 IF ⋂119894+119898119894119879119888(119875119894) = THEN

lowastAssembly tool must be changed when 119875119894is to be assembledlowast

119899119905= 119899119905+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899 ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119905

Algorithm 2 Calculation procedure for changing times of assembly tool

Step 1 Set 119894 = 0119898 = 1 119899119889= 0

Step 2 IF ⋂119894+119898119894119863119888(119875119894) = THEN

lowastAssembly direction must be changed when 119875119894is to be assembledlowast

119899119889= 119899119889+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119889

Algorithm 3 Calculation procedure for changing times of assembly direction

an assembly sequence is equal to the effect of each swap factorin a swap sequence on the assembly sequence

AP119886andAP

119887are two assembly sequences V119900119904(AP

119887ΘAP119886)

expresses the swap sequence in which AP119886is adjusted as AP

119887

It can be expressed as (5)AP119887= AP119886oplus V119900119904 (AP

119887ΘAP119886) = AP

119886oplus (V1199001 V1199002 V119900

119899)

(7)A false code of the swap sequence is shown in

Algorithm 4 For example if AP119886= (1 4 2 5 3) and AP

119887=

(2 3 5 1 4) then the swap sequence is V119900119904(AP119887ΘAP119886) =

(V1199001(1 3) V119900

2(2 5) V119900

3(3 4))

512 Frog Position Updating Strategy 119863 is the number ofswap factors contained by the moving distance matrix Dceil is the top integral function 119903119863 (119903 isin [0 1]) is the firstceil(119903119863) swap factors in the swap sequenceD For exampleif 119863 = V119900119904 = (V119900

1 V1199002 V1199003 V1199004) and 119903 = 06 then 119863 = 4

and 119903119863 = (V1199001 V1199002 V1199003)

Science and Technology of Nuclear Installations 7

Step 1 Set 119894 = 1 119895 = 0Step 2 IF AP

119886(119894) = AP

119887(119894) THEN

119895 = 119895 + 1

V = 119891119894119899119889(AP119887== AP

119886(119894))

V119900119895= V119900(119894 V)

AP119886= AP119886oplus V119900119895

ENDIFStep 3 Set 119894 = 119894 + 1

IF 119894 le 119899 minus 1 THENProceed to Step 2

ELSEProceed to Step 4

ENDIFStep 4 V119900119904 = (V119900

1 V1199002 V119900

119895)

Algorithm 4 False code of the swap sequence

Equations (1) to (3) are improved as follows

119863 = 119872(rand (0 1) lowast (119865119887Θ119865119908)) (8)

119865119908new = 119865119908old oplus 119863 (

1003817100381710038171003817119863min1003817100381710038171003817 le 119863 le

1003817100381710038171003817119863max1003817100381710038171003817) (9)

119863 = 119872(rand (0 1) lowast (119865119892Θ119865119908)) (10)

52 Diversity Control Strategy After population sorting inSFLA the grouping operator makes the best frog positionsimilar in each memeplex when the first 119898 frogs satisfy 119865

1=

1198652= sdot sdot sdot = 119865

119898(119898 is the quantity of the memeplex) Based on

(8) to (10) eachmemeplex can easily converge to the best frogposition 119865

119892of the entire population The algorithm search

space and the probability of algorithm convergence with theglobally optimal solution are reduced This study proposesa diversity control strategy to avoid homoplasy The controlpolicy is as follows

(1) Compute119873 of the preceding same frog position afterthe grouping operator

(2) If119873 lt 119898 then proceed to step (4)(3) The next population is based on standard GA(4) The population is based on the other SFLA steps

53 Improved SFLA Steps The basic steps to solve the ASPproblem by using the improved SFLA are shown in Figure 3The detailed steps are as follows

(1) Parameter initialization and population initializationfrog population size is size The number of frogmemeplex is 119898 The population iterative is iter Thelocal search iteration is mrun The maximum andminimum frog moving distances are 119863max and119863min respectively The crossover probability is pcThe adaptive mutation probability is pm

(2) Modified grouping operator the frogs are sorted indescending order according to their fitness during thepreprocessing of the grouping operator in SFLA In

Start

(3) Memetic evolution

(5) Global shuffling process

Yes

(2) Modified grouping operator

(1) Population initialization and parameter initialization

Convergence criteria are satisfied

Output the optimal solution

End

No

(4) Diversity control strategy

(local search based on swap sequence)

Figure 3 Memetic evolution procedure in SFLA

ASP a small fitness function value indicates good frogposition and good assembly sequence Therefore thegrouping operator should be modified The modifiedgrouping operator is as follows Suppose that the scaleof the frog population is 119873 which is then dividedinto 119898 memeplexes All frogs in the population arearranged in ascending order according to the fitnessfunction valueThefirst frog enters the firstmemeplexand the second frog enters the secondmemeplex andso on until the 119898 frog enters the 119898th memeplexThe (119898 + 1)th frog is then assigned back to thefirst memeplex and so on All individual frogs areassigned according to the aforementioned rule

(3) Memetic evolution (local search based on a swapsequence) in the improved SFLA memetic evolutionis modified and performed by using a local searchstrategy based on a swap sequence until the mrungeneration

(4) Optimal sampling different strategy the optimal sam-pling different strategy is included in the improvedSFLA to avoid homoplasy for each memeplex

(5) Global shuffling process the global shuffling processof the improved SFLA is similar to SFLA and updatesthe best position 119865

119892of the frog population

(6) The next step is determining whether the iterationshould be terminated according to the terminal con-dition of the algorithm If the terminal condition is

8 Science and Technology of Nuclear Installations

Table 1 Part assembly tool sets

Part number Part name Assembly tool1 Nut washer assembly 1 1198791

2 Nut washer assembly 2 1198791

3 Nut washer assembly 3 1198791

4 Nut washer assembly 4 1198791

5 Hydraulic cylinder 1198794

6 Pole 1 1198793

7 Pole 2 1198793

8 Pole 3 1198793

9 Pole 4 1198793

10 Strut 1 1198794

11 Nut 1 1198791

12 Nut 2 1198791

13 Nut 3 11987911198792

14 Bolt 1 11987911198793

15 Bolt 2 11987911198793

16 Pin 1 1198793

17 Nut 4 11987911198792

18 Pin 2 1198793

19 Central pin 1198793

20 Back plate 1198794

21 Strut 2 1198794

22 Nut washer assembly 5 1198791

23 Nut washer assembly 6 1198791

24 Nut washer assembly 7 1198791

25 Nut washer assembly 8 1198791

26 Axis 1198793

27 Hydraulic pressure scissors 1198794

28 Hydraulic pressure shear blades 1 1198795

29 Hydraulic pressure shear blades 2 1198795

satisfied then the iteration ends Otherwise return tostep (2)

6 Experiment and Analysis

The application program based on the improved SFLAis compiled under MATLAB environment The computerenvironment of the application program consists of a 20GHzCPU 2GBmemory andWindows 7 32-bit operating systemThe hydraulic pressure shear which contains 29 parts is usedfor the ASP experiment The exploded view of the hydraulicpressure shear is shown in Figure 4 The components of theassembly tool sets are listed in Table 1

61 ASP Experiment Based on SFLA After conducting anorthogonal experiment on the assembly of the hydraulicpressure shear the algorithm rapidly identifies an optimalassembly sequence when the weighting factors of the eval-uating indicator in the fitness function are 119888

119891= 4 119888

119904=

05 119888119905= 02 and 119888

119889= 03 If the memeplex has few local

search iterations then it also undergoes few evolution timeswhich reduces information exchange within the memeplex

If the memeplex has many local search iterations then itundergoes multiple local searches which increases algorithmsearch time and makes the best frog position of variousmemeplexes similar It also causes the algorithm to carry outGA several times and thus slows down the convergence rate ofthe algorithm If the maximum moving distance 119863max is toosmall then the global algorithm search capability is reducedIt causes the algorithm to easily fall into a local search If119863maxis too large then the algorithm is unable to find the globallyoptimal solutionAftermultiple comparison experiments thealgorithm optimization capability is observed to be optimalwhen the local search iteration of the memeplex is 10crossover probability is 08 adaptive mutation probabilityis 01 maximum moving distance is 8 minimum movingdistance is 1 and frog quantity in thememeplex ismaintainedat 30

The ASP experiment is conducted with population sizes60 120 180 and 240 given that each parameter value of theimproved SFLA and the weighting factor of the evaluatingindicator in the fitness function are the same The fitnessfunction value distributed the optimal assembly sequencefrom the results of the ASP experiment The analysis resultsare shown in Figure 5 in which the number of algorithmiterations is 600 and that of repeating operation times is 50A lot of experiments show that the fitness value of the globaloptimal assembly sequence 119865 is 21 As shown in Figure 5the distributed situation of the local optimal fitness value iswithin the following ranges 21 to 30 31 to 40 41 to 5051 to 60 and gt60 As shown in Figure 5 the distributedsituations of fitness value of local optimal assembly sequencediffer along with various population sizes When populationsize is 60 and time of experiment is 50 there is only onefitness value of local optimal assembly sequence in sections21 to 30 As population size increases the quantity offitness value of local optimal assembly sequence identifiedby the algorithm in this section gradually increases Whenpopulation size increases to 240 the quantity identifiedby the algorithm in this section is 20 As population sizeincrease the quantity of outstanding assembly sequenceswhose fitness value is smaller gradually increases As shownin Table 2 the increase in algorithm population size reducesalgorithm iteration efficiency and the operation time of thealgorithm is extended In this experiment the probabilityof the global optimal assembly sequence identified by thealgorithm is highest when population size is 240 and theaverage consuming time of a single experiment is in the rangeof acceptable with 515S

The mean and optimal average fitness of iteration in50 experiments when population size is 120 are shown inFigure 6 As algorithm iteration increases the optimal fitnessaverage value fluctuates However the optimal and averagevalues of the average fitness steadily decrease from the overalltendency which shows that the stability of the algorithm isgood and the algorithm convergence rate at the later periodis slow

The mean and optimal average fitness of 1 of the 50experiments in which the global optimal assembly sequenceis obtained when population size is 120 is shown in Figure 7In the algorithm implementation the mean fitness exhibits a

Science and Technology of Nuclear Installations 9

1

2

3 45

6

7

89

10

11

12

13

14

1516

1718

19

20

21

22

23

24

25

26

2728

29

Z

X

Y

Figure 4 Exploded view of the hydraulic pressure shear

huge fluctuation because the optimal sample is homoplasyAccording to the diversity control strategy the algorithmexecutes GA and increases average fitness The homoplasy ofthe algorithm improves after executing GA

62 Algorithm Comparison Experiment An algorithm com-parison experiment is conducted among improved SFLA(SFLA-GA) GA SFLA PSO and AMPSO [7] to verify theperformance of the improved SFLA for the ASP problemAMPSO ismodifiedmethod for ASP in RHM in our previouswork [7]

A hydraulic pressure shear is employed to carry out ASPwith the same programming and PC environments as thoseindicated in Section 61 Moreover the parameter 119891

119889= 21

and values of other related parameters in AMPSO are thesame as those presented in [11] Inertia weight in PSO is 06and the values of the other parameters are similar to thoseof AMPSO The experiment results with population sizes of60 and 240 are shown in Table 3 The algorithm convergencecurves of different population sizes are shown in Figures 8and 9 (ie variation of population average fitness values alongwith the iterations)

As shown in Table 3 the probability of a feasible assemblysequence identified by the improved SFLA is enhanced whenpopulation size increasesThe probability is higher than thosefor SFLA and GA but lower than those for AMPSO andPSO Under the same population size the improved SFLAexhibits a superior assembly sequence than those of GA andSFLA The value of the optimal assembly sequence fitnessfunction identified by the improved SFLA is less than those

identified by GA and SFLA When population size is 60the improved SFLA obtains an acceptable assembly sequenceand the fitness function value is 29 However GA and SFLAare unable to obtain an acceptable assembly sequence evenwhen the population size is 240 The fitness values of theoptimal assembly sequence are similar in PSO and AMPSOThe execution time of the improved SFLA is slightly less thanthat of GA and significantly less than those of AMPSO andPSO Consequently the efficiency of the proposed algorithmis acceptable Based on the local optimal fitness averagevalue of the algorithm the improved SFLA exhibits a higherconvergence rate than those of GA and SFLA and is nearthose of AMPSO and PSO The improved SFLA is evenbetter than PSO when population size is 240 Therefore theoptimization capability efficiency and convergence rate ofthe improved SFLA are better than those of GA whereasits optimization capability and convergence rate are betterthan those of SFLA The overall performance of the SFLA-GA for solving ASP problems proposed in this study issimilar to those of AMPSO and PSO As shown in Figures8 and 9 the stochastic initializing population qualities ofthe five algorithms are approximately similar Therefore thepreceding analysis is reliable

7 Conclusions

RHM is an important mean of ensuring the reliabilityof radioactive equipment and has a wide application inradioactive installations RHMP predetermine the mainte-nance procedures of radioactive equipment during the design

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Science and Technology of Nuclear Installations 3

33] studied visualization and simulation of a nuclear facilitydisassembling process and established an RHM system

In the aforementioned studies maintenance sequencewas generally obtained in an exploratory manner because ofthe lack of a guided optimized maintenance sequence Suchsequencemay be inefficient and the optimal solutionmight beomitted without intelligent support Introducing intelligentplanning technologies to ASP is essential to further enhancevirtual maintenance planning [7]

24 Intelligent ASP for Complex Products Assembling prod-ucts (by humans or by robots) is the act of combiningparts of equipment ASP obtains the order for each part andsubassembly which is then inserted into an incrementallyexpanding subassembly that eventually leads to a final assem-bly ASP is a combinatorial problem in complex productsSolving this problem with human involvement is difficultand impractical because of the combinatorial explosion issueResearch in intelligence assembly sequencing has rapidlyincreased in recent years The intelligence ASP problem isregarded as a discrete search and optimization problemVarious artificial intelligence approaches have been proposedrecently including graph theory [34 35] subassembly detec-tion [36 37] motion planning [38] and evolution algorithm[11]

Evolution algorithms provide new solutions to vari-ous complex optimization problems by imitating the self-organization mechanism of natural biological communitiesand the adaptive ability of evolution Chen and Xiao [39]developed an enhancing artificial bee colony algorithmwith self-adaptive searching strategy and artificial immunenetwork operators for global optimization Xu et al [40]developed an improved genetic algorithm for distributionnetwork planning Chen and Ju [41] developed a novelartificial bee colony algorithm for solving the supply chainnetwork design under disruption scenarios Cheng et al[42] developed a metaheuristics for airport gate assignmentLorpunmance and Sap [43] developed an ant colony opti-mization for dynamic job scheduling in grid environmentEvolution algorithms such as GA [44ndash46] the ant colonyalgorithm [47 48] the particle swarm algorithm [49ndash52] andthe artificial bee colony algorithm [53] have been studiedrecently in ASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems and hasbeen found to be effective in searching for global solutionsRahimi-Vahed andMirzaei proposed a hybrid multiobjectiveSFLA for a mixed model assembly line sequencing prob-lem [54] and bicriteria permutation flow shop schedulingproblem [55] Li et al [56] proposed an effective SFLA fora multiobjective flexible job shop scheduling problems Fangand Wang [57] introduced an effective SFLA for a resource-constrained project scheduling problem Li et al [58] devel-oped a modified SFLA for continuous optimization Theaforementioned studies show that SFLA is a simple robustand fast algorithm for solving combinatorial optimizationproblems

Start

(3) Memetic evolution (by local search)

(4) Global shuffling process

Yes

(2) Grouping operation

(1) Population initialization and parameter initialization

Convergence criterions are satisfied

Output the optimal solution

End

No

Figure 1 SFLA procedure

3 SFLA

SFLA is a metaheuristic optimization method that identifiessolutions by simulating collaboration behaviors and inter-active information similar to a frog community searchingfor food in a natural environment This algorithm dividespopulation into several subpopulations and the evolutionof memes is driven by the exchange of global informationamong subpopulations and the local evolutionary searchwithin subpopulation

SFLA is described in detail as follows The populationconsists of several frogs and each frog is a solution to theproblem The population is divided into several subpopula-tions through a grouping operator to simulate frog groupingbehaviors Each subpopulation is called a memeplex Amemeplex is composed of frogs with the same meme thatperform local searches Each frog has its own idea but isalso influenced by other frogs in the same memeplex Frogsadjust their positions through memetic evolution After apredefined number of memetic evolutionary times frogs indifferent subpopulations exchange information through aglobal shuffling process The alternating memetic evolutionand global shuffling process make the frogs jump out of thelocal optimum and evolve toward the global optimum

The SFLAprocedure is illustrated in Figure 1Thedetailedprocedure is as follows

(1) Population initialization and parameter initializationSFLA initially creates a population of 119865 frogs as acertain solution amount The 119894th frog (ie the 119894thsolution to the problem) is represented as 119865

119894=

4 Science and Technology of Nuclear Installations

(1198911198941 1198911198942 119891

119894119904) where 119904 is the solution space dimen-

sion(2) Grouping operator a grouping operator separates119865 frogs into 119898 memeplex which contains 119899 frogsaccording to their fitness order The frogs are sortedin descending order according to their fitness Thefirst frog is assigned to the first memeplex the secondfrog is assigned to the second memeplex the 119898thfrog is assigned to the 119898th memeplex the (119898 +1)th frog is assigned back to the first memeplexand so on The best and worst positions of eachfrog in each memeplex are indicated as 119865

119887and 119865

119908

respectively and the frog in the best position in theentire population is indicated as 119865

119892

(3) Memetic evolution during the memetic evolution ofthe memeplex the frog in the worst position 119865

119908is

updated through a local search in its memeplex Thenew position of the worst frog is updated accordingto (1) to (3)

119863 = 119872(rand (0 1) lowast (119865119887 minus 119865119908)) (1)

119865119908new = 119865119908old + 119863 (minus119863max le 119863 le 119863max) (2)

119863 = 119872(rand (0 1) lowast (119865119892 minus 119865119908)) (3)

where119863 is themoving distancematrix119872(rand(0 1))is the random number matrix within the range [0 1]and 119863max is the maximum distance that the frog ispermitted to moveThe memetic evolution procedure in SFLA is illus-trated in Figure 2If the new position of the worst frog is better than itsprevious position after (1) and (2) are calculated thenit replaces the position of the worst frog and the worstposition 119865

119908is recalculated Otherwise substitute 119865

119892

for 119865119887and repeat the worst position of the frog

updating calculations in (3) and (2) If the frog in theworst position still cannot obtain a better positionthen a frog with a new position is stochasticallyproduced to replace the frog in the worst position 119865

119908

Consequently each memeplex follows a predefinedmemetic evolutionary time

(4) Global shuffling process all frogs are mixed andsorted in descending order according to their fitnessThe memeplex is then divided according to the neworder and then Step (2) is repeated

(5) Local evolution and global shuffling processes con-tinue until convergence is achieved

4 ASP Problem Statement

The final goal of ASP is to enhance assembly efficiency andreduce assembly difficulties and costs Several parametersare involved to achieve this objective This study employsseveral essential parameters as evaluating indicators includ-ing geometric feasibility assembly stability changing times

Start

Yes

End

No

More than the predefined memetic evolutionary times

No

Yes

No

Yes

Find Fw Fb Fg

Update Fw by Eqs (1) (2)

New Fw is better than old Fw

New Fw is better than old Fw

Stochastically produces a new Fw

New Fw replaces Fb

Fg replaces update Fw by (3) and (2)Fb

Figure 2 Memetic evolution procedure in SFLA

of assembly tool and changing times of assembly directionThe fitness function is ultimately developed through theevaluating indicators

41 Geometric Feasibility The assembly direction is dividedinto six types of direction as follows 119889(119896) = +119909 +119910 +119911 minus119909minus119910 minus119911 Interference values 119868

119894119895119889119896(119896 = 1 2 6 and 119889

119896isin

119889(119896)) describe whether part 119875119894interfere with part 119875

119895when

moving along 119889119896direction 119868

119894119895119889119896is as follows

119868119894119895119889119896

=

0 119901119894does not interfere with 119901

119895in the 119889

119896direction

1 119901119894interferes with 119901

119895in the 119889

119896direction

(4)

Suppose that AP = (1198751 1198752 119875

119899) is an assembly

sequence The part set AP1= 1198751 1198752 119875

119894minus1 is the set in

which parts have been assembled and 119875119894is the part to be

assembled Then 119878119896(119875119894) (119896 = 1 2 6) is the sum of the

Science and Technology of Nuclear Installations 5

IF 119888119875119894119875119895= 2 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSEIF 119888119875119894119875119895== 0 119895 isin [1 119894 minus 1] THEN

Assembly operation is unstableELSE

IF 119904119875119895119875119894= 1 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSE IF 119904

119875119895119875119894== 0 THEN

Assembly operation is unstableENDIF

ENDIF

Algorithm 1 Stability evaluation method of the assemblysequence

interference values which is 119875119894and each part in AP

1when 119875

119894

is assembled along 119889119896direction 119878

119896(119875119894) is as follows

119878119896(119875119894) =

119894minus1

sum

119895=1

119868119875119894119875119895119889119896 (5)

If 119878119896(119875119894) = 0 then 119875

119894can be assembled along 119889

119896

Otherwise 119875119894cannot be assembled along 119889

119896 In this case we

obtain119863119862(119875119894) = 119889

119896| 119878119896(119875119894) = 0 that is the feasible assembly

direction set of 119875119894 For each 119875

119894(1 lt 119894 le 119899) if 119863119862(119875

119894) = 120601

then AP is the feasible assembly sequence otherwise AP isthe infeasible assembly sequence 119899

119891is expressed as the total

times of assembly interference of APThe value of 119899119891is equal

to the total times of119863119862(119875119894) = 120601 (1 lt 119894 le 119899) in AP

42 Assembly Stabilities In the actual assembly process partsmay become unstable because of gravity Several assemblyoperationsmust use a jig or auxiliary tool tomaintain stabilitywhen a part is unstable during the assembly process whichresults in an inefficient assembly Therefore the stability ofthe assembly sequence should be evaluated

The augmented adjacencymatrix119862 = (119888119894119895)119899times119899

and supportmatrix 119878 = (119904

119894119895)119899times119899

are defined to evaluate the stability of theassembly sequence In the augmented adjacency matrix 119888

119894119895

expresses the connection type between 119875119894and 119875119895 For a stable

connection 119888119894119895= 2 for a contact connection 119888

119894119895= 1 and

for a noncontact connection 119888119894119895= 0 In the support matrix

119904119894119895expresses the support type between 119875

119894and 119875

119895 For stable

support 119904119894119895= 1 otherwise 119904

119894119895= 0

Suppose that AP = (1199011 1199012 119901

119899) is an assembly

sequence The part set AP1= 1199011 1199012 119901

119894minus1 is expressed

as the parts having been assembled and 119875119894is expressed as the

part to be assembled The stability evaluation method of theassembly sequence is shown in Algorithm 1 In this study 119899

119904

expresses the times of the assembly sequence stable operationA smaller 119899

119904indicates a more stable assembly sequence

43 Changing Times of Assembly Tool Given the particularityof each assembly part different assembly tools should be usedin the actual assembly process Changing the assembly tool

leads to a long assembly time and high cost for the assemblyprocessTherefore changing times of assembly tool should beas few as possible

Suppose the assembly sequence is AP = (1198751 1198752 119875

119899)

and assembly tool sequence of AP is 119879119888 119879119888(119875119894) is expressed

as the assembly tool of 119875119894 Assembly tool of each part is

determined by the characteristic of each part and the availableassembly tool The assembly tool sequence for an AP as wellas the optimal assembly tool sequence is predeterminedChanging times of assembly tool 119899

119905are calculated as shown

in Algorithm 2

44 Changing Times of Assembly Direction The reducedchanging times of assembly direction shorten assembly timeand enhance assembly efficiency Supposing the assemblysequence is AP = (119875

1 1198752 119875

119899) changing times of assembly

direction 119899119889are calculated as shown in Algorithm 3

45 Fitness Function Different radioactive equipment undervarious environmentsmay have varying influence degrees forthe evaluating indicators Therefore weighting factors mustbe determined according to the actual situation A penaltyfunction 119888

119891119899119891is applied to infeasible assembly sequence to

speed up the algorithm convergence rate Then the weightedfitness function is as follows

119891 = 119888119904119899119904+ 119888119905119899119905+ 119888119889119899119889+ 119888119891119899119891 (6)

where 119888119904 119888119905 119888119889 and 119888

119891are the weighting factors for each

evaluating indicator and 119888119891must be generally larger than the

other three weighting factors (ie 119888119891ge (1198992)max119888

119904 119888119905 119888119889)

In this study a small fitness function value indicates goodposition of the frog and good assembly sequence

5 Improved SFLA for ASP

51 Local Search Strategy Based on a Swap Sequence ASP is acombinatorial optimization problem in which each solutiondimension is discrete A GA can solve the discrete optimiza-tion problem by using crossover andmutation operatorsTheimproved SFLA introduces a local search strategy based on aswap sequence to address this problem

511 Swap Factor and Swap Sequence

(1) Swap Factor Suppose that an assembly sequence thatincludes 119899 parts is expressed as AP = (119875

1 1198752 119875

119899) The

function of swap factor V119900(120574 120596) is to swap the positions of120574 and 120596 to form a new assembly sequence For example if theinitial assembly sequence is AP = (2 4 3 5 1) and the swapfactor is V119900 = V119900(2 4) then AP1015840 = AP oplus V119900 = (2 5 3 4 1)oplus indicates that the swap factor is acting on the assemblysequence

(2) Swap Sequence V119900119904 = (V1199001 V1199002 V119900

119899) expresses a

swap sequence that consists of 119899 swap factors in whichV1199001 V1199002 V119900

119899are the swap factors and their order does not

satisfy the commutative lawThe effect of a swap sequence on

6 Science and Technology of Nuclear Installations

Step 1 Set 119894 = 0119898 = 1 119899119905= 0

Step 2 IF ⋂119894+119898119894119879119888(119875119894) = THEN

lowastAssembly tool must be changed when 119875119894is to be assembledlowast

119899119905= 119899119905+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899 ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119905

Algorithm 2 Calculation procedure for changing times of assembly tool

Step 1 Set 119894 = 0119898 = 1 119899119889= 0

Step 2 IF ⋂119894+119898119894119863119888(119875119894) = THEN

lowastAssembly direction must be changed when 119875119894is to be assembledlowast

119899119889= 119899119889+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119889

Algorithm 3 Calculation procedure for changing times of assembly direction

an assembly sequence is equal to the effect of each swap factorin a swap sequence on the assembly sequence

AP119886andAP

119887are two assembly sequences V119900119904(AP

119887ΘAP119886)

expresses the swap sequence in which AP119886is adjusted as AP

119887

It can be expressed as (5)AP119887= AP119886oplus V119900119904 (AP

119887ΘAP119886) = AP

119886oplus (V1199001 V1199002 V119900

119899)

(7)A false code of the swap sequence is shown in

Algorithm 4 For example if AP119886= (1 4 2 5 3) and AP

119887=

(2 3 5 1 4) then the swap sequence is V119900119904(AP119887ΘAP119886) =

(V1199001(1 3) V119900

2(2 5) V119900

3(3 4))

512 Frog Position Updating Strategy 119863 is the number ofswap factors contained by the moving distance matrix Dceil is the top integral function 119903119863 (119903 isin [0 1]) is the firstceil(119903119863) swap factors in the swap sequenceD For exampleif 119863 = V119900119904 = (V119900

1 V1199002 V1199003 V1199004) and 119903 = 06 then 119863 = 4

and 119903119863 = (V1199001 V1199002 V1199003)

Science and Technology of Nuclear Installations 7

Step 1 Set 119894 = 1 119895 = 0Step 2 IF AP

119886(119894) = AP

119887(119894) THEN

119895 = 119895 + 1

V = 119891119894119899119889(AP119887== AP

119886(119894))

V119900119895= V119900(119894 V)

AP119886= AP119886oplus V119900119895

ENDIFStep 3 Set 119894 = 119894 + 1

IF 119894 le 119899 minus 1 THENProceed to Step 2

ELSEProceed to Step 4

ENDIFStep 4 V119900119904 = (V119900

1 V1199002 V119900

119895)

Algorithm 4 False code of the swap sequence

Equations (1) to (3) are improved as follows

119863 = 119872(rand (0 1) lowast (119865119887Θ119865119908)) (8)

119865119908new = 119865119908old oplus 119863 (

1003817100381710038171003817119863min1003817100381710038171003817 le 119863 le

1003817100381710038171003817119863max1003817100381710038171003817) (9)

119863 = 119872(rand (0 1) lowast (119865119892Θ119865119908)) (10)

52 Diversity Control Strategy After population sorting inSFLA the grouping operator makes the best frog positionsimilar in each memeplex when the first 119898 frogs satisfy 119865

1=

1198652= sdot sdot sdot = 119865

119898(119898 is the quantity of the memeplex) Based on

(8) to (10) eachmemeplex can easily converge to the best frogposition 119865

119892of the entire population The algorithm search

space and the probability of algorithm convergence with theglobally optimal solution are reduced This study proposesa diversity control strategy to avoid homoplasy The controlpolicy is as follows

(1) Compute119873 of the preceding same frog position afterthe grouping operator

(2) If119873 lt 119898 then proceed to step (4)(3) The next population is based on standard GA(4) The population is based on the other SFLA steps

53 Improved SFLA Steps The basic steps to solve the ASPproblem by using the improved SFLA are shown in Figure 3The detailed steps are as follows

(1) Parameter initialization and population initializationfrog population size is size The number of frogmemeplex is 119898 The population iterative is iter Thelocal search iteration is mrun The maximum andminimum frog moving distances are 119863max and119863min respectively The crossover probability is pcThe adaptive mutation probability is pm

(2) Modified grouping operator the frogs are sorted indescending order according to their fitness during thepreprocessing of the grouping operator in SFLA In

Start

(3) Memetic evolution

(5) Global shuffling process

Yes

(2) Modified grouping operator

(1) Population initialization and parameter initialization

Convergence criteria are satisfied

Output the optimal solution

End

No

(4) Diversity control strategy

(local search based on swap sequence)

Figure 3 Memetic evolution procedure in SFLA

ASP a small fitness function value indicates good frogposition and good assembly sequence Therefore thegrouping operator should be modified The modifiedgrouping operator is as follows Suppose that the scaleof the frog population is 119873 which is then dividedinto 119898 memeplexes All frogs in the population arearranged in ascending order according to the fitnessfunction valueThefirst frog enters the firstmemeplexand the second frog enters the secondmemeplex andso on until the 119898 frog enters the 119898th memeplexThe (119898 + 1)th frog is then assigned back to thefirst memeplex and so on All individual frogs areassigned according to the aforementioned rule

(3) Memetic evolution (local search based on a swapsequence) in the improved SFLA memetic evolutionis modified and performed by using a local searchstrategy based on a swap sequence until the mrungeneration

(4) Optimal sampling different strategy the optimal sam-pling different strategy is included in the improvedSFLA to avoid homoplasy for each memeplex

(5) Global shuffling process the global shuffling processof the improved SFLA is similar to SFLA and updatesthe best position 119865

119892of the frog population

(6) The next step is determining whether the iterationshould be terminated according to the terminal con-dition of the algorithm If the terminal condition is

8 Science and Technology of Nuclear Installations

Table 1 Part assembly tool sets

Part number Part name Assembly tool1 Nut washer assembly 1 1198791

2 Nut washer assembly 2 1198791

3 Nut washer assembly 3 1198791

4 Nut washer assembly 4 1198791

5 Hydraulic cylinder 1198794

6 Pole 1 1198793

7 Pole 2 1198793

8 Pole 3 1198793

9 Pole 4 1198793

10 Strut 1 1198794

11 Nut 1 1198791

12 Nut 2 1198791

13 Nut 3 11987911198792

14 Bolt 1 11987911198793

15 Bolt 2 11987911198793

16 Pin 1 1198793

17 Nut 4 11987911198792

18 Pin 2 1198793

19 Central pin 1198793

20 Back plate 1198794

21 Strut 2 1198794

22 Nut washer assembly 5 1198791

23 Nut washer assembly 6 1198791

24 Nut washer assembly 7 1198791

25 Nut washer assembly 8 1198791

26 Axis 1198793

27 Hydraulic pressure scissors 1198794

28 Hydraulic pressure shear blades 1 1198795

29 Hydraulic pressure shear blades 2 1198795

satisfied then the iteration ends Otherwise return tostep (2)

6 Experiment and Analysis

The application program based on the improved SFLAis compiled under MATLAB environment The computerenvironment of the application program consists of a 20GHzCPU 2GBmemory andWindows 7 32-bit operating systemThe hydraulic pressure shear which contains 29 parts is usedfor the ASP experiment The exploded view of the hydraulicpressure shear is shown in Figure 4 The components of theassembly tool sets are listed in Table 1

61 ASP Experiment Based on SFLA After conducting anorthogonal experiment on the assembly of the hydraulicpressure shear the algorithm rapidly identifies an optimalassembly sequence when the weighting factors of the eval-uating indicator in the fitness function are 119888

119891= 4 119888

119904=

05 119888119905= 02 and 119888

119889= 03 If the memeplex has few local

search iterations then it also undergoes few evolution timeswhich reduces information exchange within the memeplex

If the memeplex has many local search iterations then itundergoes multiple local searches which increases algorithmsearch time and makes the best frog position of variousmemeplexes similar It also causes the algorithm to carry outGA several times and thus slows down the convergence rate ofthe algorithm If the maximum moving distance 119863max is toosmall then the global algorithm search capability is reducedIt causes the algorithm to easily fall into a local search If119863maxis too large then the algorithm is unable to find the globallyoptimal solutionAftermultiple comparison experiments thealgorithm optimization capability is observed to be optimalwhen the local search iteration of the memeplex is 10crossover probability is 08 adaptive mutation probabilityis 01 maximum moving distance is 8 minimum movingdistance is 1 and frog quantity in thememeplex ismaintainedat 30

The ASP experiment is conducted with population sizes60 120 180 and 240 given that each parameter value of theimproved SFLA and the weighting factor of the evaluatingindicator in the fitness function are the same The fitnessfunction value distributed the optimal assembly sequencefrom the results of the ASP experiment The analysis resultsare shown in Figure 5 in which the number of algorithmiterations is 600 and that of repeating operation times is 50A lot of experiments show that the fitness value of the globaloptimal assembly sequence 119865 is 21 As shown in Figure 5the distributed situation of the local optimal fitness value iswithin the following ranges 21 to 30 31 to 40 41 to 5051 to 60 and gt60 As shown in Figure 5 the distributedsituations of fitness value of local optimal assembly sequencediffer along with various population sizes When populationsize is 60 and time of experiment is 50 there is only onefitness value of local optimal assembly sequence in sections21 to 30 As population size increases the quantity offitness value of local optimal assembly sequence identifiedby the algorithm in this section gradually increases Whenpopulation size increases to 240 the quantity identifiedby the algorithm in this section is 20 As population sizeincrease the quantity of outstanding assembly sequenceswhose fitness value is smaller gradually increases As shownin Table 2 the increase in algorithm population size reducesalgorithm iteration efficiency and the operation time of thealgorithm is extended In this experiment the probabilityof the global optimal assembly sequence identified by thealgorithm is highest when population size is 240 and theaverage consuming time of a single experiment is in the rangeof acceptable with 515S

The mean and optimal average fitness of iteration in50 experiments when population size is 120 are shown inFigure 6 As algorithm iteration increases the optimal fitnessaverage value fluctuates However the optimal and averagevalues of the average fitness steadily decrease from the overalltendency which shows that the stability of the algorithm isgood and the algorithm convergence rate at the later periodis slow

The mean and optimal average fitness of 1 of the 50experiments in which the global optimal assembly sequenceis obtained when population size is 120 is shown in Figure 7In the algorithm implementation the mean fitness exhibits a

Science and Technology of Nuclear Installations 9

1

2

3 45

6

7

89

10

11

12

13

14

1516

1718

19

20

21

22

23

24

25

26

2728

29

Z

X

Y

Figure 4 Exploded view of the hydraulic pressure shear

huge fluctuation because the optimal sample is homoplasyAccording to the diversity control strategy the algorithmexecutes GA and increases average fitness The homoplasy ofthe algorithm improves after executing GA

62 Algorithm Comparison Experiment An algorithm com-parison experiment is conducted among improved SFLA(SFLA-GA) GA SFLA PSO and AMPSO [7] to verify theperformance of the improved SFLA for the ASP problemAMPSO ismodifiedmethod for ASP in RHM in our previouswork [7]

A hydraulic pressure shear is employed to carry out ASPwith the same programming and PC environments as thoseindicated in Section 61 Moreover the parameter 119891

119889= 21

and values of other related parameters in AMPSO are thesame as those presented in [11] Inertia weight in PSO is 06and the values of the other parameters are similar to thoseof AMPSO The experiment results with population sizes of60 and 240 are shown in Table 3 The algorithm convergencecurves of different population sizes are shown in Figures 8and 9 (ie variation of population average fitness values alongwith the iterations)

As shown in Table 3 the probability of a feasible assemblysequence identified by the improved SFLA is enhanced whenpopulation size increasesThe probability is higher than thosefor SFLA and GA but lower than those for AMPSO andPSO Under the same population size the improved SFLAexhibits a superior assembly sequence than those of GA andSFLA The value of the optimal assembly sequence fitnessfunction identified by the improved SFLA is less than those

identified by GA and SFLA When population size is 60the improved SFLA obtains an acceptable assembly sequenceand the fitness function value is 29 However GA and SFLAare unable to obtain an acceptable assembly sequence evenwhen the population size is 240 The fitness values of theoptimal assembly sequence are similar in PSO and AMPSOThe execution time of the improved SFLA is slightly less thanthat of GA and significantly less than those of AMPSO andPSO Consequently the efficiency of the proposed algorithmis acceptable Based on the local optimal fitness averagevalue of the algorithm the improved SFLA exhibits a higherconvergence rate than those of GA and SFLA and is nearthose of AMPSO and PSO The improved SFLA is evenbetter than PSO when population size is 240 Therefore theoptimization capability efficiency and convergence rate ofthe improved SFLA are better than those of GA whereasits optimization capability and convergence rate are betterthan those of SFLA The overall performance of the SFLA-GA for solving ASP problems proposed in this study issimilar to those of AMPSO and PSO As shown in Figures8 and 9 the stochastic initializing population qualities ofthe five algorithms are approximately similar Therefore thepreceding analysis is reliable

7 Conclusions

RHM is an important mean of ensuring the reliabilityof radioactive equipment and has a wide application inradioactive installations RHMP predetermine the mainte-nance procedures of radioactive equipment during the design

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

4 Science and Technology of Nuclear Installations

(1198911198941 1198911198942 119891

119894119904) where 119904 is the solution space dimen-

sion(2) Grouping operator a grouping operator separates119865 frogs into 119898 memeplex which contains 119899 frogsaccording to their fitness order The frogs are sortedin descending order according to their fitness Thefirst frog is assigned to the first memeplex the secondfrog is assigned to the second memeplex the 119898thfrog is assigned to the 119898th memeplex the (119898 +1)th frog is assigned back to the first memeplexand so on The best and worst positions of eachfrog in each memeplex are indicated as 119865

119887and 119865

119908

respectively and the frog in the best position in theentire population is indicated as 119865

119892

(3) Memetic evolution during the memetic evolution ofthe memeplex the frog in the worst position 119865

119908is

updated through a local search in its memeplex Thenew position of the worst frog is updated accordingto (1) to (3)

119863 = 119872(rand (0 1) lowast (119865119887 minus 119865119908)) (1)

119865119908new = 119865119908old + 119863 (minus119863max le 119863 le 119863max) (2)

119863 = 119872(rand (0 1) lowast (119865119892 minus 119865119908)) (3)

where119863 is themoving distancematrix119872(rand(0 1))is the random number matrix within the range [0 1]and 119863max is the maximum distance that the frog ispermitted to moveThe memetic evolution procedure in SFLA is illus-trated in Figure 2If the new position of the worst frog is better than itsprevious position after (1) and (2) are calculated thenit replaces the position of the worst frog and the worstposition 119865

119908is recalculated Otherwise substitute 119865

119892

for 119865119887and repeat the worst position of the frog

updating calculations in (3) and (2) If the frog in theworst position still cannot obtain a better positionthen a frog with a new position is stochasticallyproduced to replace the frog in the worst position 119865

119908

Consequently each memeplex follows a predefinedmemetic evolutionary time

(4) Global shuffling process all frogs are mixed andsorted in descending order according to their fitnessThe memeplex is then divided according to the neworder and then Step (2) is repeated

(5) Local evolution and global shuffling processes con-tinue until convergence is achieved

4 ASP Problem Statement

The final goal of ASP is to enhance assembly efficiency andreduce assembly difficulties and costs Several parametersare involved to achieve this objective This study employsseveral essential parameters as evaluating indicators includ-ing geometric feasibility assembly stability changing times

Start

Yes

End

No

More than the predefined memetic evolutionary times

No

Yes

No

Yes

Find Fw Fb Fg

Update Fw by Eqs (1) (2)

New Fw is better than old Fw

New Fw is better than old Fw

Stochastically produces a new Fw

New Fw replaces Fb

Fg replaces update Fw by (3) and (2)Fb

Figure 2 Memetic evolution procedure in SFLA

of assembly tool and changing times of assembly directionThe fitness function is ultimately developed through theevaluating indicators

41 Geometric Feasibility The assembly direction is dividedinto six types of direction as follows 119889(119896) = +119909 +119910 +119911 minus119909minus119910 minus119911 Interference values 119868

119894119895119889119896(119896 = 1 2 6 and 119889

119896isin

119889(119896)) describe whether part 119875119894interfere with part 119875

119895when

moving along 119889119896direction 119868

119894119895119889119896is as follows

119868119894119895119889119896

=

0 119901119894does not interfere with 119901

119895in the 119889

119896direction

1 119901119894interferes with 119901

119895in the 119889

119896direction

(4)

Suppose that AP = (1198751 1198752 119875

119899) is an assembly

sequence The part set AP1= 1198751 1198752 119875

119894minus1 is the set in

which parts have been assembled and 119875119894is the part to be

assembled Then 119878119896(119875119894) (119896 = 1 2 6) is the sum of the

Science and Technology of Nuclear Installations 5

IF 119888119875119894119875119895= 2 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSEIF 119888119875119894119875119895== 0 119895 isin [1 119894 minus 1] THEN

Assembly operation is unstableELSE

IF 119904119875119895119875119894= 1 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSE IF 119904

119875119895119875119894== 0 THEN

Assembly operation is unstableENDIF

ENDIF

Algorithm 1 Stability evaluation method of the assemblysequence

interference values which is 119875119894and each part in AP

1when 119875

119894

is assembled along 119889119896direction 119878

119896(119875119894) is as follows

119878119896(119875119894) =

119894minus1

sum

119895=1

119868119875119894119875119895119889119896 (5)

If 119878119896(119875119894) = 0 then 119875

119894can be assembled along 119889

119896

Otherwise 119875119894cannot be assembled along 119889

119896 In this case we

obtain119863119862(119875119894) = 119889

119896| 119878119896(119875119894) = 0 that is the feasible assembly

direction set of 119875119894 For each 119875

119894(1 lt 119894 le 119899) if 119863119862(119875

119894) = 120601

then AP is the feasible assembly sequence otherwise AP isthe infeasible assembly sequence 119899

119891is expressed as the total

times of assembly interference of APThe value of 119899119891is equal

to the total times of119863119862(119875119894) = 120601 (1 lt 119894 le 119899) in AP

42 Assembly Stabilities In the actual assembly process partsmay become unstable because of gravity Several assemblyoperationsmust use a jig or auxiliary tool tomaintain stabilitywhen a part is unstable during the assembly process whichresults in an inefficient assembly Therefore the stability ofthe assembly sequence should be evaluated

The augmented adjacencymatrix119862 = (119888119894119895)119899times119899

and supportmatrix 119878 = (119904

119894119895)119899times119899

are defined to evaluate the stability of theassembly sequence In the augmented adjacency matrix 119888

119894119895

expresses the connection type between 119875119894and 119875119895 For a stable

connection 119888119894119895= 2 for a contact connection 119888

119894119895= 1 and

for a noncontact connection 119888119894119895= 0 In the support matrix

119904119894119895expresses the support type between 119875

119894and 119875

119895 For stable

support 119904119894119895= 1 otherwise 119904

119894119895= 0

Suppose that AP = (1199011 1199012 119901

119899) is an assembly

sequence The part set AP1= 1199011 1199012 119901

119894minus1 is expressed

as the parts having been assembled and 119875119894is expressed as the

part to be assembled The stability evaluation method of theassembly sequence is shown in Algorithm 1 In this study 119899

119904

expresses the times of the assembly sequence stable operationA smaller 119899

119904indicates a more stable assembly sequence

43 Changing Times of Assembly Tool Given the particularityof each assembly part different assembly tools should be usedin the actual assembly process Changing the assembly tool

leads to a long assembly time and high cost for the assemblyprocessTherefore changing times of assembly tool should beas few as possible

Suppose the assembly sequence is AP = (1198751 1198752 119875

119899)

and assembly tool sequence of AP is 119879119888 119879119888(119875119894) is expressed

as the assembly tool of 119875119894 Assembly tool of each part is

determined by the characteristic of each part and the availableassembly tool The assembly tool sequence for an AP as wellas the optimal assembly tool sequence is predeterminedChanging times of assembly tool 119899

119905are calculated as shown

in Algorithm 2

44 Changing Times of Assembly Direction The reducedchanging times of assembly direction shorten assembly timeand enhance assembly efficiency Supposing the assemblysequence is AP = (119875

1 1198752 119875

119899) changing times of assembly

direction 119899119889are calculated as shown in Algorithm 3

45 Fitness Function Different radioactive equipment undervarious environmentsmay have varying influence degrees forthe evaluating indicators Therefore weighting factors mustbe determined according to the actual situation A penaltyfunction 119888

119891119899119891is applied to infeasible assembly sequence to

speed up the algorithm convergence rate Then the weightedfitness function is as follows

119891 = 119888119904119899119904+ 119888119905119899119905+ 119888119889119899119889+ 119888119891119899119891 (6)

where 119888119904 119888119905 119888119889 and 119888

119891are the weighting factors for each

evaluating indicator and 119888119891must be generally larger than the

other three weighting factors (ie 119888119891ge (1198992)max119888

119904 119888119905 119888119889)

In this study a small fitness function value indicates goodposition of the frog and good assembly sequence

5 Improved SFLA for ASP

51 Local Search Strategy Based on a Swap Sequence ASP is acombinatorial optimization problem in which each solutiondimension is discrete A GA can solve the discrete optimiza-tion problem by using crossover andmutation operatorsTheimproved SFLA introduces a local search strategy based on aswap sequence to address this problem

511 Swap Factor and Swap Sequence

(1) Swap Factor Suppose that an assembly sequence thatincludes 119899 parts is expressed as AP = (119875

1 1198752 119875

119899) The

function of swap factor V119900(120574 120596) is to swap the positions of120574 and 120596 to form a new assembly sequence For example if theinitial assembly sequence is AP = (2 4 3 5 1) and the swapfactor is V119900 = V119900(2 4) then AP1015840 = AP oplus V119900 = (2 5 3 4 1)oplus indicates that the swap factor is acting on the assemblysequence

(2) Swap Sequence V119900119904 = (V1199001 V1199002 V119900

119899) expresses a

swap sequence that consists of 119899 swap factors in whichV1199001 V1199002 V119900

119899are the swap factors and their order does not

satisfy the commutative lawThe effect of a swap sequence on

6 Science and Technology of Nuclear Installations

Step 1 Set 119894 = 0119898 = 1 119899119905= 0

Step 2 IF ⋂119894+119898119894119879119888(119875119894) = THEN

lowastAssembly tool must be changed when 119875119894is to be assembledlowast

119899119905= 119899119905+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899 ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119905

Algorithm 2 Calculation procedure for changing times of assembly tool

Step 1 Set 119894 = 0119898 = 1 119899119889= 0

Step 2 IF ⋂119894+119898119894119863119888(119875119894) = THEN

lowastAssembly direction must be changed when 119875119894is to be assembledlowast

119899119889= 119899119889+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119889

Algorithm 3 Calculation procedure for changing times of assembly direction

an assembly sequence is equal to the effect of each swap factorin a swap sequence on the assembly sequence

AP119886andAP

119887are two assembly sequences V119900119904(AP

119887ΘAP119886)

expresses the swap sequence in which AP119886is adjusted as AP

119887

It can be expressed as (5)AP119887= AP119886oplus V119900119904 (AP

119887ΘAP119886) = AP

119886oplus (V1199001 V1199002 V119900

119899)

(7)A false code of the swap sequence is shown in

Algorithm 4 For example if AP119886= (1 4 2 5 3) and AP

119887=

(2 3 5 1 4) then the swap sequence is V119900119904(AP119887ΘAP119886) =

(V1199001(1 3) V119900

2(2 5) V119900

3(3 4))

512 Frog Position Updating Strategy 119863 is the number ofswap factors contained by the moving distance matrix Dceil is the top integral function 119903119863 (119903 isin [0 1]) is the firstceil(119903119863) swap factors in the swap sequenceD For exampleif 119863 = V119900119904 = (V119900

1 V1199002 V1199003 V1199004) and 119903 = 06 then 119863 = 4

and 119903119863 = (V1199001 V1199002 V1199003)

Science and Technology of Nuclear Installations 7

Step 1 Set 119894 = 1 119895 = 0Step 2 IF AP

119886(119894) = AP

119887(119894) THEN

119895 = 119895 + 1

V = 119891119894119899119889(AP119887== AP

119886(119894))

V119900119895= V119900(119894 V)

AP119886= AP119886oplus V119900119895

ENDIFStep 3 Set 119894 = 119894 + 1

IF 119894 le 119899 minus 1 THENProceed to Step 2

ELSEProceed to Step 4

ENDIFStep 4 V119900119904 = (V119900

1 V1199002 V119900

119895)

Algorithm 4 False code of the swap sequence

Equations (1) to (3) are improved as follows

119863 = 119872(rand (0 1) lowast (119865119887Θ119865119908)) (8)

119865119908new = 119865119908old oplus 119863 (

1003817100381710038171003817119863min1003817100381710038171003817 le 119863 le

1003817100381710038171003817119863max1003817100381710038171003817) (9)

119863 = 119872(rand (0 1) lowast (119865119892Θ119865119908)) (10)

52 Diversity Control Strategy After population sorting inSFLA the grouping operator makes the best frog positionsimilar in each memeplex when the first 119898 frogs satisfy 119865

1=

1198652= sdot sdot sdot = 119865

119898(119898 is the quantity of the memeplex) Based on

(8) to (10) eachmemeplex can easily converge to the best frogposition 119865

119892of the entire population The algorithm search

space and the probability of algorithm convergence with theglobally optimal solution are reduced This study proposesa diversity control strategy to avoid homoplasy The controlpolicy is as follows

(1) Compute119873 of the preceding same frog position afterthe grouping operator

(2) If119873 lt 119898 then proceed to step (4)(3) The next population is based on standard GA(4) The population is based on the other SFLA steps

53 Improved SFLA Steps The basic steps to solve the ASPproblem by using the improved SFLA are shown in Figure 3The detailed steps are as follows

(1) Parameter initialization and population initializationfrog population size is size The number of frogmemeplex is 119898 The population iterative is iter Thelocal search iteration is mrun The maximum andminimum frog moving distances are 119863max and119863min respectively The crossover probability is pcThe adaptive mutation probability is pm

(2) Modified grouping operator the frogs are sorted indescending order according to their fitness during thepreprocessing of the grouping operator in SFLA In

Start

(3) Memetic evolution

(5) Global shuffling process

Yes

(2) Modified grouping operator

(1) Population initialization and parameter initialization

Convergence criteria are satisfied

Output the optimal solution

End

No

(4) Diversity control strategy

(local search based on swap sequence)

Figure 3 Memetic evolution procedure in SFLA

ASP a small fitness function value indicates good frogposition and good assembly sequence Therefore thegrouping operator should be modified The modifiedgrouping operator is as follows Suppose that the scaleof the frog population is 119873 which is then dividedinto 119898 memeplexes All frogs in the population arearranged in ascending order according to the fitnessfunction valueThefirst frog enters the firstmemeplexand the second frog enters the secondmemeplex andso on until the 119898 frog enters the 119898th memeplexThe (119898 + 1)th frog is then assigned back to thefirst memeplex and so on All individual frogs areassigned according to the aforementioned rule

(3) Memetic evolution (local search based on a swapsequence) in the improved SFLA memetic evolutionis modified and performed by using a local searchstrategy based on a swap sequence until the mrungeneration

(4) Optimal sampling different strategy the optimal sam-pling different strategy is included in the improvedSFLA to avoid homoplasy for each memeplex

(5) Global shuffling process the global shuffling processof the improved SFLA is similar to SFLA and updatesthe best position 119865

119892of the frog population

(6) The next step is determining whether the iterationshould be terminated according to the terminal con-dition of the algorithm If the terminal condition is

8 Science and Technology of Nuclear Installations

Table 1 Part assembly tool sets

Part number Part name Assembly tool1 Nut washer assembly 1 1198791

2 Nut washer assembly 2 1198791

3 Nut washer assembly 3 1198791

4 Nut washer assembly 4 1198791

5 Hydraulic cylinder 1198794

6 Pole 1 1198793

7 Pole 2 1198793

8 Pole 3 1198793

9 Pole 4 1198793

10 Strut 1 1198794

11 Nut 1 1198791

12 Nut 2 1198791

13 Nut 3 11987911198792

14 Bolt 1 11987911198793

15 Bolt 2 11987911198793

16 Pin 1 1198793

17 Nut 4 11987911198792

18 Pin 2 1198793

19 Central pin 1198793

20 Back plate 1198794

21 Strut 2 1198794

22 Nut washer assembly 5 1198791

23 Nut washer assembly 6 1198791

24 Nut washer assembly 7 1198791

25 Nut washer assembly 8 1198791

26 Axis 1198793

27 Hydraulic pressure scissors 1198794

28 Hydraulic pressure shear blades 1 1198795

29 Hydraulic pressure shear blades 2 1198795

satisfied then the iteration ends Otherwise return tostep (2)

6 Experiment and Analysis

The application program based on the improved SFLAis compiled under MATLAB environment The computerenvironment of the application program consists of a 20GHzCPU 2GBmemory andWindows 7 32-bit operating systemThe hydraulic pressure shear which contains 29 parts is usedfor the ASP experiment The exploded view of the hydraulicpressure shear is shown in Figure 4 The components of theassembly tool sets are listed in Table 1

61 ASP Experiment Based on SFLA After conducting anorthogonal experiment on the assembly of the hydraulicpressure shear the algorithm rapidly identifies an optimalassembly sequence when the weighting factors of the eval-uating indicator in the fitness function are 119888

119891= 4 119888

119904=

05 119888119905= 02 and 119888

119889= 03 If the memeplex has few local

search iterations then it also undergoes few evolution timeswhich reduces information exchange within the memeplex

If the memeplex has many local search iterations then itundergoes multiple local searches which increases algorithmsearch time and makes the best frog position of variousmemeplexes similar It also causes the algorithm to carry outGA several times and thus slows down the convergence rate ofthe algorithm If the maximum moving distance 119863max is toosmall then the global algorithm search capability is reducedIt causes the algorithm to easily fall into a local search If119863maxis too large then the algorithm is unable to find the globallyoptimal solutionAftermultiple comparison experiments thealgorithm optimization capability is observed to be optimalwhen the local search iteration of the memeplex is 10crossover probability is 08 adaptive mutation probabilityis 01 maximum moving distance is 8 minimum movingdistance is 1 and frog quantity in thememeplex ismaintainedat 30

The ASP experiment is conducted with population sizes60 120 180 and 240 given that each parameter value of theimproved SFLA and the weighting factor of the evaluatingindicator in the fitness function are the same The fitnessfunction value distributed the optimal assembly sequencefrom the results of the ASP experiment The analysis resultsare shown in Figure 5 in which the number of algorithmiterations is 600 and that of repeating operation times is 50A lot of experiments show that the fitness value of the globaloptimal assembly sequence 119865 is 21 As shown in Figure 5the distributed situation of the local optimal fitness value iswithin the following ranges 21 to 30 31 to 40 41 to 5051 to 60 and gt60 As shown in Figure 5 the distributedsituations of fitness value of local optimal assembly sequencediffer along with various population sizes When populationsize is 60 and time of experiment is 50 there is only onefitness value of local optimal assembly sequence in sections21 to 30 As population size increases the quantity offitness value of local optimal assembly sequence identifiedby the algorithm in this section gradually increases Whenpopulation size increases to 240 the quantity identifiedby the algorithm in this section is 20 As population sizeincrease the quantity of outstanding assembly sequenceswhose fitness value is smaller gradually increases As shownin Table 2 the increase in algorithm population size reducesalgorithm iteration efficiency and the operation time of thealgorithm is extended In this experiment the probabilityof the global optimal assembly sequence identified by thealgorithm is highest when population size is 240 and theaverage consuming time of a single experiment is in the rangeof acceptable with 515S

The mean and optimal average fitness of iteration in50 experiments when population size is 120 are shown inFigure 6 As algorithm iteration increases the optimal fitnessaverage value fluctuates However the optimal and averagevalues of the average fitness steadily decrease from the overalltendency which shows that the stability of the algorithm isgood and the algorithm convergence rate at the later periodis slow

The mean and optimal average fitness of 1 of the 50experiments in which the global optimal assembly sequenceis obtained when population size is 120 is shown in Figure 7In the algorithm implementation the mean fitness exhibits a

Science and Technology of Nuclear Installations 9

1

2

3 45

6

7

89

10

11

12

13

14

1516

1718

19

20

21

22

23

24

25

26

2728

29

Z

X

Y

Figure 4 Exploded view of the hydraulic pressure shear

huge fluctuation because the optimal sample is homoplasyAccording to the diversity control strategy the algorithmexecutes GA and increases average fitness The homoplasy ofthe algorithm improves after executing GA

62 Algorithm Comparison Experiment An algorithm com-parison experiment is conducted among improved SFLA(SFLA-GA) GA SFLA PSO and AMPSO [7] to verify theperformance of the improved SFLA for the ASP problemAMPSO ismodifiedmethod for ASP in RHM in our previouswork [7]

A hydraulic pressure shear is employed to carry out ASPwith the same programming and PC environments as thoseindicated in Section 61 Moreover the parameter 119891

119889= 21

and values of other related parameters in AMPSO are thesame as those presented in [11] Inertia weight in PSO is 06and the values of the other parameters are similar to thoseof AMPSO The experiment results with population sizes of60 and 240 are shown in Table 3 The algorithm convergencecurves of different population sizes are shown in Figures 8and 9 (ie variation of population average fitness values alongwith the iterations)

As shown in Table 3 the probability of a feasible assemblysequence identified by the improved SFLA is enhanced whenpopulation size increasesThe probability is higher than thosefor SFLA and GA but lower than those for AMPSO andPSO Under the same population size the improved SFLAexhibits a superior assembly sequence than those of GA andSFLA The value of the optimal assembly sequence fitnessfunction identified by the improved SFLA is less than those

identified by GA and SFLA When population size is 60the improved SFLA obtains an acceptable assembly sequenceand the fitness function value is 29 However GA and SFLAare unable to obtain an acceptable assembly sequence evenwhen the population size is 240 The fitness values of theoptimal assembly sequence are similar in PSO and AMPSOThe execution time of the improved SFLA is slightly less thanthat of GA and significantly less than those of AMPSO andPSO Consequently the efficiency of the proposed algorithmis acceptable Based on the local optimal fitness averagevalue of the algorithm the improved SFLA exhibits a higherconvergence rate than those of GA and SFLA and is nearthose of AMPSO and PSO The improved SFLA is evenbetter than PSO when population size is 240 Therefore theoptimization capability efficiency and convergence rate ofthe improved SFLA are better than those of GA whereasits optimization capability and convergence rate are betterthan those of SFLA The overall performance of the SFLA-GA for solving ASP problems proposed in this study issimilar to those of AMPSO and PSO As shown in Figures8 and 9 the stochastic initializing population qualities ofthe five algorithms are approximately similar Therefore thepreceding analysis is reliable

7 Conclusions

RHM is an important mean of ensuring the reliabilityof radioactive equipment and has a wide application inradioactive installations RHMP predetermine the mainte-nance procedures of radioactive equipment during the design

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Science and Technology of Nuclear Installations 5

IF 119888119875119894119875119895= 2 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSEIF 119888119875119894119875119895== 0 119895 isin [1 119894 minus 1] THEN

Assembly operation is unstableELSE

IF 119904119875119895119875119894= 1 119895 isin [1 119894 minus 1] exists THEN

Assembly operation is stableELSE IF 119904

119875119895119875119894== 0 THEN

Assembly operation is unstableENDIF

ENDIF

Algorithm 1 Stability evaluation method of the assemblysequence

interference values which is 119875119894and each part in AP

1when 119875

119894

is assembled along 119889119896direction 119878

119896(119875119894) is as follows

119878119896(119875119894) =

119894minus1

sum

119895=1

119868119875119894119875119895119889119896 (5)

If 119878119896(119875119894) = 0 then 119875

119894can be assembled along 119889

119896

Otherwise 119875119894cannot be assembled along 119889

119896 In this case we

obtain119863119862(119875119894) = 119889

119896| 119878119896(119875119894) = 0 that is the feasible assembly

direction set of 119875119894 For each 119875

119894(1 lt 119894 le 119899) if 119863119862(119875

119894) = 120601

then AP is the feasible assembly sequence otherwise AP isthe infeasible assembly sequence 119899

119891is expressed as the total

times of assembly interference of APThe value of 119899119891is equal

to the total times of119863119862(119875119894) = 120601 (1 lt 119894 le 119899) in AP

42 Assembly Stabilities In the actual assembly process partsmay become unstable because of gravity Several assemblyoperationsmust use a jig or auxiliary tool tomaintain stabilitywhen a part is unstable during the assembly process whichresults in an inefficient assembly Therefore the stability ofthe assembly sequence should be evaluated

The augmented adjacencymatrix119862 = (119888119894119895)119899times119899

and supportmatrix 119878 = (119904

119894119895)119899times119899

are defined to evaluate the stability of theassembly sequence In the augmented adjacency matrix 119888

119894119895

expresses the connection type between 119875119894and 119875119895 For a stable

connection 119888119894119895= 2 for a contact connection 119888

119894119895= 1 and

for a noncontact connection 119888119894119895= 0 In the support matrix

119904119894119895expresses the support type between 119875

119894and 119875

119895 For stable

support 119904119894119895= 1 otherwise 119904

119894119895= 0

Suppose that AP = (1199011 1199012 119901

119899) is an assembly

sequence The part set AP1= 1199011 1199012 119901

119894minus1 is expressed

as the parts having been assembled and 119875119894is expressed as the

part to be assembled The stability evaluation method of theassembly sequence is shown in Algorithm 1 In this study 119899

119904

expresses the times of the assembly sequence stable operationA smaller 119899

119904indicates a more stable assembly sequence

43 Changing Times of Assembly Tool Given the particularityof each assembly part different assembly tools should be usedin the actual assembly process Changing the assembly tool

leads to a long assembly time and high cost for the assemblyprocessTherefore changing times of assembly tool should beas few as possible

Suppose the assembly sequence is AP = (1198751 1198752 119875

119899)

and assembly tool sequence of AP is 119879119888 119879119888(119875119894) is expressed

as the assembly tool of 119875119894 Assembly tool of each part is

determined by the characteristic of each part and the availableassembly tool The assembly tool sequence for an AP as wellas the optimal assembly tool sequence is predeterminedChanging times of assembly tool 119899

119905are calculated as shown

in Algorithm 2

44 Changing Times of Assembly Direction The reducedchanging times of assembly direction shorten assembly timeand enhance assembly efficiency Supposing the assemblysequence is AP = (119875

1 1198752 119875

119899) changing times of assembly

direction 119899119889are calculated as shown in Algorithm 3

45 Fitness Function Different radioactive equipment undervarious environmentsmay have varying influence degrees forthe evaluating indicators Therefore weighting factors mustbe determined according to the actual situation A penaltyfunction 119888

119891119899119891is applied to infeasible assembly sequence to

speed up the algorithm convergence rate Then the weightedfitness function is as follows

119891 = 119888119904119899119904+ 119888119905119899119905+ 119888119889119899119889+ 119888119891119899119891 (6)

where 119888119904 119888119905 119888119889 and 119888

119891are the weighting factors for each

evaluating indicator and 119888119891must be generally larger than the

other three weighting factors (ie 119888119891ge (1198992)max119888

119904 119888119905 119888119889)

In this study a small fitness function value indicates goodposition of the frog and good assembly sequence

5 Improved SFLA for ASP

51 Local Search Strategy Based on a Swap Sequence ASP is acombinatorial optimization problem in which each solutiondimension is discrete A GA can solve the discrete optimiza-tion problem by using crossover andmutation operatorsTheimproved SFLA introduces a local search strategy based on aswap sequence to address this problem

511 Swap Factor and Swap Sequence

(1) Swap Factor Suppose that an assembly sequence thatincludes 119899 parts is expressed as AP = (119875

1 1198752 119875

119899) The

function of swap factor V119900(120574 120596) is to swap the positions of120574 and 120596 to form a new assembly sequence For example if theinitial assembly sequence is AP = (2 4 3 5 1) and the swapfactor is V119900 = V119900(2 4) then AP1015840 = AP oplus V119900 = (2 5 3 4 1)oplus indicates that the swap factor is acting on the assemblysequence

(2) Swap Sequence V119900119904 = (V1199001 V1199002 V119900

119899) expresses a

swap sequence that consists of 119899 swap factors in whichV1199001 V1199002 V119900

119899are the swap factors and their order does not

satisfy the commutative lawThe effect of a swap sequence on

6 Science and Technology of Nuclear Installations

Step 1 Set 119894 = 0119898 = 1 119899119905= 0

Step 2 IF ⋂119894+119898119894119879119888(119875119894) = THEN

lowastAssembly tool must be changed when 119875119894is to be assembledlowast

119899119905= 119899119905+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899 ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119905

Algorithm 2 Calculation procedure for changing times of assembly tool

Step 1 Set 119894 = 0119898 = 1 119899119889= 0

Step 2 IF ⋂119894+119898119894119863119888(119875119894) = THEN

lowastAssembly direction must be changed when 119875119894is to be assembledlowast

119899119889= 119899119889+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119889

Algorithm 3 Calculation procedure for changing times of assembly direction

an assembly sequence is equal to the effect of each swap factorin a swap sequence on the assembly sequence

AP119886andAP

119887are two assembly sequences V119900119904(AP

119887ΘAP119886)

expresses the swap sequence in which AP119886is adjusted as AP

119887

It can be expressed as (5)AP119887= AP119886oplus V119900119904 (AP

119887ΘAP119886) = AP

119886oplus (V1199001 V1199002 V119900

119899)

(7)A false code of the swap sequence is shown in

Algorithm 4 For example if AP119886= (1 4 2 5 3) and AP

119887=

(2 3 5 1 4) then the swap sequence is V119900119904(AP119887ΘAP119886) =

(V1199001(1 3) V119900

2(2 5) V119900

3(3 4))

512 Frog Position Updating Strategy 119863 is the number ofswap factors contained by the moving distance matrix Dceil is the top integral function 119903119863 (119903 isin [0 1]) is the firstceil(119903119863) swap factors in the swap sequenceD For exampleif 119863 = V119900119904 = (V119900

1 V1199002 V1199003 V1199004) and 119903 = 06 then 119863 = 4

and 119903119863 = (V1199001 V1199002 V1199003)

Science and Technology of Nuclear Installations 7

Step 1 Set 119894 = 1 119895 = 0Step 2 IF AP

119886(119894) = AP

119887(119894) THEN

119895 = 119895 + 1

V = 119891119894119899119889(AP119887== AP

119886(119894))

V119900119895= V119900(119894 V)

AP119886= AP119886oplus V119900119895

ENDIFStep 3 Set 119894 = 119894 + 1

IF 119894 le 119899 minus 1 THENProceed to Step 2

ELSEProceed to Step 4

ENDIFStep 4 V119900119904 = (V119900

1 V1199002 V119900

119895)

Algorithm 4 False code of the swap sequence

Equations (1) to (3) are improved as follows

119863 = 119872(rand (0 1) lowast (119865119887Θ119865119908)) (8)

119865119908new = 119865119908old oplus 119863 (

1003817100381710038171003817119863min1003817100381710038171003817 le 119863 le

1003817100381710038171003817119863max1003817100381710038171003817) (9)

119863 = 119872(rand (0 1) lowast (119865119892Θ119865119908)) (10)

52 Diversity Control Strategy After population sorting inSFLA the grouping operator makes the best frog positionsimilar in each memeplex when the first 119898 frogs satisfy 119865

1=

1198652= sdot sdot sdot = 119865

119898(119898 is the quantity of the memeplex) Based on

(8) to (10) eachmemeplex can easily converge to the best frogposition 119865

119892of the entire population The algorithm search

space and the probability of algorithm convergence with theglobally optimal solution are reduced This study proposesa diversity control strategy to avoid homoplasy The controlpolicy is as follows

(1) Compute119873 of the preceding same frog position afterthe grouping operator

(2) If119873 lt 119898 then proceed to step (4)(3) The next population is based on standard GA(4) The population is based on the other SFLA steps

53 Improved SFLA Steps The basic steps to solve the ASPproblem by using the improved SFLA are shown in Figure 3The detailed steps are as follows

(1) Parameter initialization and population initializationfrog population size is size The number of frogmemeplex is 119898 The population iterative is iter Thelocal search iteration is mrun The maximum andminimum frog moving distances are 119863max and119863min respectively The crossover probability is pcThe adaptive mutation probability is pm

(2) Modified grouping operator the frogs are sorted indescending order according to their fitness during thepreprocessing of the grouping operator in SFLA In

Start

(3) Memetic evolution

(5) Global shuffling process

Yes

(2) Modified grouping operator

(1) Population initialization and parameter initialization

Convergence criteria are satisfied

Output the optimal solution

End

No

(4) Diversity control strategy

(local search based on swap sequence)

Figure 3 Memetic evolution procedure in SFLA

ASP a small fitness function value indicates good frogposition and good assembly sequence Therefore thegrouping operator should be modified The modifiedgrouping operator is as follows Suppose that the scaleof the frog population is 119873 which is then dividedinto 119898 memeplexes All frogs in the population arearranged in ascending order according to the fitnessfunction valueThefirst frog enters the firstmemeplexand the second frog enters the secondmemeplex andso on until the 119898 frog enters the 119898th memeplexThe (119898 + 1)th frog is then assigned back to thefirst memeplex and so on All individual frogs areassigned according to the aforementioned rule

(3) Memetic evolution (local search based on a swapsequence) in the improved SFLA memetic evolutionis modified and performed by using a local searchstrategy based on a swap sequence until the mrungeneration

(4) Optimal sampling different strategy the optimal sam-pling different strategy is included in the improvedSFLA to avoid homoplasy for each memeplex

(5) Global shuffling process the global shuffling processof the improved SFLA is similar to SFLA and updatesthe best position 119865

119892of the frog population

(6) The next step is determining whether the iterationshould be terminated according to the terminal con-dition of the algorithm If the terminal condition is

8 Science and Technology of Nuclear Installations

Table 1 Part assembly tool sets

Part number Part name Assembly tool1 Nut washer assembly 1 1198791

2 Nut washer assembly 2 1198791

3 Nut washer assembly 3 1198791

4 Nut washer assembly 4 1198791

5 Hydraulic cylinder 1198794

6 Pole 1 1198793

7 Pole 2 1198793

8 Pole 3 1198793

9 Pole 4 1198793

10 Strut 1 1198794

11 Nut 1 1198791

12 Nut 2 1198791

13 Nut 3 11987911198792

14 Bolt 1 11987911198793

15 Bolt 2 11987911198793

16 Pin 1 1198793

17 Nut 4 11987911198792

18 Pin 2 1198793

19 Central pin 1198793

20 Back plate 1198794

21 Strut 2 1198794

22 Nut washer assembly 5 1198791

23 Nut washer assembly 6 1198791

24 Nut washer assembly 7 1198791

25 Nut washer assembly 8 1198791

26 Axis 1198793

27 Hydraulic pressure scissors 1198794

28 Hydraulic pressure shear blades 1 1198795

29 Hydraulic pressure shear blades 2 1198795

satisfied then the iteration ends Otherwise return tostep (2)

6 Experiment and Analysis

The application program based on the improved SFLAis compiled under MATLAB environment The computerenvironment of the application program consists of a 20GHzCPU 2GBmemory andWindows 7 32-bit operating systemThe hydraulic pressure shear which contains 29 parts is usedfor the ASP experiment The exploded view of the hydraulicpressure shear is shown in Figure 4 The components of theassembly tool sets are listed in Table 1

61 ASP Experiment Based on SFLA After conducting anorthogonal experiment on the assembly of the hydraulicpressure shear the algorithm rapidly identifies an optimalassembly sequence when the weighting factors of the eval-uating indicator in the fitness function are 119888

119891= 4 119888

119904=

05 119888119905= 02 and 119888

119889= 03 If the memeplex has few local

search iterations then it also undergoes few evolution timeswhich reduces information exchange within the memeplex

If the memeplex has many local search iterations then itundergoes multiple local searches which increases algorithmsearch time and makes the best frog position of variousmemeplexes similar It also causes the algorithm to carry outGA several times and thus slows down the convergence rate ofthe algorithm If the maximum moving distance 119863max is toosmall then the global algorithm search capability is reducedIt causes the algorithm to easily fall into a local search If119863maxis too large then the algorithm is unable to find the globallyoptimal solutionAftermultiple comparison experiments thealgorithm optimization capability is observed to be optimalwhen the local search iteration of the memeplex is 10crossover probability is 08 adaptive mutation probabilityis 01 maximum moving distance is 8 minimum movingdistance is 1 and frog quantity in thememeplex ismaintainedat 30

The ASP experiment is conducted with population sizes60 120 180 and 240 given that each parameter value of theimproved SFLA and the weighting factor of the evaluatingindicator in the fitness function are the same The fitnessfunction value distributed the optimal assembly sequencefrom the results of the ASP experiment The analysis resultsare shown in Figure 5 in which the number of algorithmiterations is 600 and that of repeating operation times is 50A lot of experiments show that the fitness value of the globaloptimal assembly sequence 119865 is 21 As shown in Figure 5the distributed situation of the local optimal fitness value iswithin the following ranges 21 to 30 31 to 40 41 to 5051 to 60 and gt60 As shown in Figure 5 the distributedsituations of fitness value of local optimal assembly sequencediffer along with various population sizes When populationsize is 60 and time of experiment is 50 there is only onefitness value of local optimal assembly sequence in sections21 to 30 As population size increases the quantity offitness value of local optimal assembly sequence identifiedby the algorithm in this section gradually increases Whenpopulation size increases to 240 the quantity identifiedby the algorithm in this section is 20 As population sizeincrease the quantity of outstanding assembly sequenceswhose fitness value is smaller gradually increases As shownin Table 2 the increase in algorithm population size reducesalgorithm iteration efficiency and the operation time of thealgorithm is extended In this experiment the probabilityof the global optimal assembly sequence identified by thealgorithm is highest when population size is 240 and theaverage consuming time of a single experiment is in the rangeof acceptable with 515S

The mean and optimal average fitness of iteration in50 experiments when population size is 120 are shown inFigure 6 As algorithm iteration increases the optimal fitnessaverage value fluctuates However the optimal and averagevalues of the average fitness steadily decrease from the overalltendency which shows that the stability of the algorithm isgood and the algorithm convergence rate at the later periodis slow

The mean and optimal average fitness of 1 of the 50experiments in which the global optimal assembly sequenceis obtained when population size is 120 is shown in Figure 7In the algorithm implementation the mean fitness exhibits a

Science and Technology of Nuclear Installations 9

1

2

3 45

6

7

89

10

11

12

13

14

1516

1718

19

20

21

22

23

24

25

26

2728

29

Z

X

Y

Figure 4 Exploded view of the hydraulic pressure shear

huge fluctuation because the optimal sample is homoplasyAccording to the diversity control strategy the algorithmexecutes GA and increases average fitness The homoplasy ofthe algorithm improves after executing GA

62 Algorithm Comparison Experiment An algorithm com-parison experiment is conducted among improved SFLA(SFLA-GA) GA SFLA PSO and AMPSO [7] to verify theperformance of the improved SFLA for the ASP problemAMPSO ismodifiedmethod for ASP in RHM in our previouswork [7]

A hydraulic pressure shear is employed to carry out ASPwith the same programming and PC environments as thoseindicated in Section 61 Moreover the parameter 119891

119889= 21

and values of other related parameters in AMPSO are thesame as those presented in [11] Inertia weight in PSO is 06and the values of the other parameters are similar to thoseof AMPSO The experiment results with population sizes of60 and 240 are shown in Table 3 The algorithm convergencecurves of different population sizes are shown in Figures 8and 9 (ie variation of population average fitness values alongwith the iterations)

As shown in Table 3 the probability of a feasible assemblysequence identified by the improved SFLA is enhanced whenpopulation size increasesThe probability is higher than thosefor SFLA and GA but lower than those for AMPSO andPSO Under the same population size the improved SFLAexhibits a superior assembly sequence than those of GA andSFLA The value of the optimal assembly sequence fitnessfunction identified by the improved SFLA is less than those

identified by GA and SFLA When population size is 60the improved SFLA obtains an acceptable assembly sequenceand the fitness function value is 29 However GA and SFLAare unable to obtain an acceptable assembly sequence evenwhen the population size is 240 The fitness values of theoptimal assembly sequence are similar in PSO and AMPSOThe execution time of the improved SFLA is slightly less thanthat of GA and significantly less than those of AMPSO andPSO Consequently the efficiency of the proposed algorithmis acceptable Based on the local optimal fitness averagevalue of the algorithm the improved SFLA exhibits a higherconvergence rate than those of GA and SFLA and is nearthose of AMPSO and PSO The improved SFLA is evenbetter than PSO when population size is 240 Therefore theoptimization capability efficiency and convergence rate ofthe improved SFLA are better than those of GA whereasits optimization capability and convergence rate are betterthan those of SFLA The overall performance of the SFLA-GA for solving ASP problems proposed in this study issimilar to those of AMPSO and PSO As shown in Figures8 and 9 the stochastic initializing population qualities ofthe five algorithms are approximately similar Therefore thepreceding analysis is reliable

7 Conclusions

RHM is an important mean of ensuring the reliabilityof radioactive equipment and has a wide application inradioactive installations RHMP predetermine the mainte-nance procedures of radioactive equipment during the design

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

6 Science and Technology of Nuclear Installations

Step 1 Set 119894 = 0119898 = 1 119899119905= 0

Step 2 IF ⋂119894+119898119894119879119888(119875119894) = THEN

lowastAssembly tool must be changed when 119875119894is to be assembledlowast

119899119905= 119899119905+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899 ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119905

Algorithm 2 Calculation procedure for changing times of assembly tool

Step 1 Set 119894 = 0119898 = 1 119899119889= 0

Step 2 IF ⋂119894+119898119894119863119888(119875119894) = THEN

lowastAssembly direction must be changed when 119875119894is to be assembledlowast

119899119889= 119899119889+ 1

Proceed to Step 4ELSE

Proceed to Step 3ENDIF

Step 3 Set 119898 = 119898 + 1IF 119894 + 119898 le 119899 THEN

Proceed to Step 2ELSE

Proceed to Step 5ENDIF

Step 4 IF 119894 + 119898 = 119899ThenProceed to Step 5

ELSESet 119894 = 119894 + 119898119898 = 1Proceed to Step 2

ENDIFStep 5 Output 119899

119889

Algorithm 3 Calculation procedure for changing times of assembly direction

an assembly sequence is equal to the effect of each swap factorin a swap sequence on the assembly sequence

AP119886andAP

119887are two assembly sequences V119900119904(AP

119887ΘAP119886)

expresses the swap sequence in which AP119886is adjusted as AP

119887

It can be expressed as (5)AP119887= AP119886oplus V119900119904 (AP

119887ΘAP119886) = AP

119886oplus (V1199001 V1199002 V119900

119899)

(7)A false code of the swap sequence is shown in

Algorithm 4 For example if AP119886= (1 4 2 5 3) and AP

119887=

(2 3 5 1 4) then the swap sequence is V119900119904(AP119887ΘAP119886) =

(V1199001(1 3) V119900

2(2 5) V119900

3(3 4))

512 Frog Position Updating Strategy 119863 is the number ofswap factors contained by the moving distance matrix Dceil is the top integral function 119903119863 (119903 isin [0 1]) is the firstceil(119903119863) swap factors in the swap sequenceD For exampleif 119863 = V119900119904 = (V119900

1 V1199002 V1199003 V1199004) and 119903 = 06 then 119863 = 4

and 119903119863 = (V1199001 V1199002 V1199003)

Science and Technology of Nuclear Installations 7

Step 1 Set 119894 = 1 119895 = 0Step 2 IF AP

119886(119894) = AP

119887(119894) THEN

119895 = 119895 + 1

V = 119891119894119899119889(AP119887== AP

119886(119894))

V119900119895= V119900(119894 V)

AP119886= AP119886oplus V119900119895

ENDIFStep 3 Set 119894 = 119894 + 1

IF 119894 le 119899 minus 1 THENProceed to Step 2

ELSEProceed to Step 4

ENDIFStep 4 V119900119904 = (V119900

1 V1199002 V119900

119895)

Algorithm 4 False code of the swap sequence

Equations (1) to (3) are improved as follows

119863 = 119872(rand (0 1) lowast (119865119887Θ119865119908)) (8)

119865119908new = 119865119908old oplus 119863 (

1003817100381710038171003817119863min1003817100381710038171003817 le 119863 le

1003817100381710038171003817119863max1003817100381710038171003817) (9)

119863 = 119872(rand (0 1) lowast (119865119892Θ119865119908)) (10)

52 Diversity Control Strategy After population sorting inSFLA the grouping operator makes the best frog positionsimilar in each memeplex when the first 119898 frogs satisfy 119865

1=

1198652= sdot sdot sdot = 119865

119898(119898 is the quantity of the memeplex) Based on

(8) to (10) eachmemeplex can easily converge to the best frogposition 119865

119892of the entire population The algorithm search

space and the probability of algorithm convergence with theglobally optimal solution are reduced This study proposesa diversity control strategy to avoid homoplasy The controlpolicy is as follows

(1) Compute119873 of the preceding same frog position afterthe grouping operator

(2) If119873 lt 119898 then proceed to step (4)(3) The next population is based on standard GA(4) The population is based on the other SFLA steps

53 Improved SFLA Steps The basic steps to solve the ASPproblem by using the improved SFLA are shown in Figure 3The detailed steps are as follows

(1) Parameter initialization and population initializationfrog population size is size The number of frogmemeplex is 119898 The population iterative is iter Thelocal search iteration is mrun The maximum andminimum frog moving distances are 119863max and119863min respectively The crossover probability is pcThe adaptive mutation probability is pm

(2) Modified grouping operator the frogs are sorted indescending order according to their fitness during thepreprocessing of the grouping operator in SFLA In

Start

(3) Memetic evolution

(5) Global shuffling process

Yes

(2) Modified grouping operator

(1) Population initialization and parameter initialization

Convergence criteria are satisfied

Output the optimal solution

End

No

(4) Diversity control strategy

(local search based on swap sequence)

Figure 3 Memetic evolution procedure in SFLA

ASP a small fitness function value indicates good frogposition and good assembly sequence Therefore thegrouping operator should be modified The modifiedgrouping operator is as follows Suppose that the scaleof the frog population is 119873 which is then dividedinto 119898 memeplexes All frogs in the population arearranged in ascending order according to the fitnessfunction valueThefirst frog enters the firstmemeplexand the second frog enters the secondmemeplex andso on until the 119898 frog enters the 119898th memeplexThe (119898 + 1)th frog is then assigned back to thefirst memeplex and so on All individual frogs areassigned according to the aforementioned rule

(3) Memetic evolution (local search based on a swapsequence) in the improved SFLA memetic evolutionis modified and performed by using a local searchstrategy based on a swap sequence until the mrungeneration

(4) Optimal sampling different strategy the optimal sam-pling different strategy is included in the improvedSFLA to avoid homoplasy for each memeplex

(5) Global shuffling process the global shuffling processof the improved SFLA is similar to SFLA and updatesthe best position 119865

119892of the frog population

(6) The next step is determining whether the iterationshould be terminated according to the terminal con-dition of the algorithm If the terminal condition is

8 Science and Technology of Nuclear Installations

Table 1 Part assembly tool sets

Part number Part name Assembly tool1 Nut washer assembly 1 1198791

2 Nut washer assembly 2 1198791

3 Nut washer assembly 3 1198791

4 Nut washer assembly 4 1198791

5 Hydraulic cylinder 1198794

6 Pole 1 1198793

7 Pole 2 1198793

8 Pole 3 1198793

9 Pole 4 1198793

10 Strut 1 1198794

11 Nut 1 1198791

12 Nut 2 1198791

13 Nut 3 11987911198792

14 Bolt 1 11987911198793

15 Bolt 2 11987911198793

16 Pin 1 1198793

17 Nut 4 11987911198792

18 Pin 2 1198793

19 Central pin 1198793

20 Back plate 1198794

21 Strut 2 1198794

22 Nut washer assembly 5 1198791

23 Nut washer assembly 6 1198791

24 Nut washer assembly 7 1198791

25 Nut washer assembly 8 1198791

26 Axis 1198793

27 Hydraulic pressure scissors 1198794

28 Hydraulic pressure shear blades 1 1198795

29 Hydraulic pressure shear blades 2 1198795

satisfied then the iteration ends Otherwise return tostep (2)

6 Experiment and Analysis

The application program based on the improved SFLAis compiled under MATLAB environment The computerenvironment of the application program consists of a 20GHzCPU 2GBmemory andWindows 7 32-bit operating systemThe hydraulic pressure shear which contains 29 parts is usedfor the ASP experiment The exploded view of the hydraulicpressure shear is shown in Figure 4 The components of theassembly tool sets are listed in Table 1

61 ASP Experiment Based on SFLA After conducting anorthogonal experiment on the assembly of the hydraulicpressure shear the algorithm rapidly identifies an optimalassembly sequence when the weighting factors of the eval-uating indicator in the fitness function are 119888

119891= 4 119888

119904=

05 119888119905= 02 and 119888

119889= 03 If the memeplex has few local

search iterations then it also undergoes few evolution timeswhich reduces information exchange within the memeplex

If the memeplex has many local search iterations then itundergoes multiple local searches which increases algorithmsearch time and makes the best frog position of variousmemeplexes similar It also causes the algorithm to carry outGA several times and thus slows down the convergence rate ofthe algorithm If the maximum moving distance 119863max is toosmall then the global algorithm search capability is reducedIt causes the algorithm to easily fall into a local search If119863maxis too large then the algorithm is unable to find the globallyoptimal solutionAftermultiple comparison experiments thealgorithm optimization capability is observed to be optimalwhen the local search iteration of the memeplex is 10crossover probability is 08 adaptive mutation probabilityis 01 maximum moving distance is 8 minimum movingdistance is 1 and frog quantity in thememeplex ismaintainedat 30

The ASP experiment is conducted with population sizes60 120 180 and 240 given that each parameter value of theimproved SFLA and the weighting factor of the evaluatingindicator in the fitness function are the same The fitnessfunction value distributed the optimal assembly sequencefrom the results of the ASP experiment The analysis resultsare shown in Figure 5 in which the number of algorithmiterations is 600 and that of repeating operation times is 50A lot of experiments show that the fitness value of the globaloptimal assembly sequence 119865 is 21 As shown in Figure 5the distributed situation of the local optimal fitness value iswithin the following ranges 21 to 30 31 to 40 41 to 5051 to 60 and gt60 As shown in Figure 5 the distributedsituations of fitness value of local optimal assembly sequencediffer along with various population sizes When populationsize is 60 and time of experiment is 50 there is only onefitness value of local optimal assembly sequence in sections21 to 30 As population size increases the quantity offitness value of local optimal assembly sequence identifiedby the algorithm in this section gradually increases Whenpopulation size increases to 240 the quantity identifiedby the algorithm in this section is 20 As population sizeincrease the quantity of outstanding assembly sequenceswhose fitness value is smaller gradually increases As shownin Table 2 the increase in algorithm population size reducesalgorithm iteration efficiency and the operation time of thealgorithm is extended In this experiment the probabilityof the global optimal assembly sequence identified by thealgorithm is highest when population size is 240 and theaverage consuming time of a single experiment is in the rangeof acceptable with 515S

The mean and optimal average fitness of iteration in50 experiments when population size is 120 are shown inFigure 6 As algorithm iteration increases the optimal fitnessaverage value fluctuates However the optimal and averagevalues of the average fitness steadily decrease from the overalltendency which shows that the stability of the algorithm isgood and the algorithm convergence rate at the later periodis slow

The mean and optimal average fitness of 1 of the 50experiments in which the global optimal assembly sequenceis obtained when population size is 120 is shown in Figure 7In the algorithm implementation the mean fitness exhibits a

Science and Technology of Nuclear Installations 9

1

2

3 45

6

7

89

10

11

12

13

14

1516

1718

19

20

21

22

23

24

25

26

2728

29

Z

X

Y

Figure 4 Exploded view of the hydraulic pressure shear

huge fluctuation because the optimal sample is homoplasyAccording to the diversity control strategy the algorithmexecutes GA and increases average fitness The homoplasy ofthe algorithm improves after executing GA

62 Algorithm Comparison Experiment An algorithm com-parison experiment is conducted among improved SFLA(SFLA-GA) GA SFLA PSO and AMPSO [7] to verify theperformance of the improved SFLA for the ASP problemAMPSO ismodifiedmethod for ASP in RHM in our previouswork [7]

A hydraulic pressure shear is employed to carry out ASPwith the same programming and PC environments as thoseindicated in Section 61 Moreover the parameter 119891

119889= 21

and values of other related parameters in AMPSO are thesame as those presented in [11] Inertia weight in PSO is 06and the values of the other parameters are similar to thoseof AMPSO The experiment results with population sizes of60 and 240 are shown in Table 3 The algorithm convergencecurves of different population sizes are shown in Figures 8and 9 (ie variation of population average fitness values alongwith the iterations)

As shown in Table 3 the probability of a feasible assemblysequence identified by the improved SFLA is enhanced whenpopulation size increasesThe probability is higher than thosefor SFLA and GA but lower than those for AMPSO andPSO Under the same population size the improved SFLAexhibits a superior assembly sequence than those of GA andSFLA The value of the optimal assembly sequence fitnessfunction identified by the improved SFLA is less than those

identified by GA and SFLA When population size is 60the improved SFLA obtains an acceptable assembly sequenceand the fitness function value is 29 However GA and SFLAare unable to obtain an acceptable assembly sequence evenwhen the population size is 240 The fitness values of theoptimal assembly sequence are similar in PSO and AMPSOThe execution time of the improved SFLA is slightly less thanthat of GA and significantly less than those of AMPSO andPSO Consequently the efficiency of the proposed algorithmis acceptable Based on the local optimal fitness averagevalue of the algorithm the improved SFLA exhibits a higherconvergence rate than those of GA and SFLA and is nearthose of AMPSO and PSO The improved SFLA is evenbetter than PSO when population size is 240 Therefore theoptimization capability efficiency and convergence rate ofthe improved SFLA are better than those of GA whereasits optimization capability and convergence rate are betterthan those of SFLA The overall performance of the SFLA-GA for solving ASP problems proposed in this study issimilar to those of AMPSO and PSO As shown in Figures8 and 9 the stochastic initializing population qualities ofthe five algorithms are approximately similar Therefore thepreceding analysis is reliable

7 Conclusions

RHM is an important mean of ensuring the reliabilityof radioactive equipment and has a wide application inradioactive installations RHMP predetermine the mainte-nance procedures of radioactive equipment during the design

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Science and Technology of Nuclear Installations 7

Step 1 Set 119894 = 1 119895 = 0Step 2 IF AP

119886(119894) = AP

119887(119894) THEN

119895 = 119895 + 1

V = 119891119894119899119889(AP119887== AP

119886(119894))

V119900119895= V119900(119894 V)

AP119886= AP119886oplus V119900119895

ENDIFStep 3 Set 119894 = 119894 + 1

IF 119894 le 119899 minus 1 THENProceed to Step 2

ELSEProceed to Step 4

ENDIFStep 4 V119900119904 = (V119900

1 V1199002 V119900

119895)

Algorithm 4 False code of the swap sequence

Equations (1) to (3) are improved as follows

119863 = 119872(rand (0 1) lowast (119865119887Θ119865119908)) (8)

119865119908new = 119865119908old oplus 119863 (

1003817100381710038171003817119863min1003817100381710038171003817 le 119863 le

1003817100381710038171003817119863max1003817100381710038171003817) (9)

119863 = 119872(rand (0 1) lowast (119865119892Θ119865119908)) (10)

52 Diversity Control Strategy After population sorting inSFLA the grouping operator makes the best frog positionsimilar in each memeplex when the first 119898 frogs satisfy 119865

1=

1198652= sdot sdot sdot = 119865

119898(119898 is the quantity of the memeplex) Based on

(8) to (10) eachmemeplex can easily converge to the best frogposition 119865

119892of the entire population The algorithm search

space and the probability of algorithm convergence with theglobally optimal solution are reduced This study proposesa diversity control strategy to avoid homoplasy The controlpolicy is as follows

(1) Compute119873 of the preceding same frog position afterthe grouping operator

(2) If119873 lt 119898 then proceed to step (4)(3) The next population is based on standard GA(4) The population is based on the other SFLA steps

53 Improved SFLA Steps The basic steps to solve the ASPproblem by using the improved SFLA are shown in Figure 3The detailed steps are as follows

(1) Parameter initialization and population initializationfrog population size is size The number of frogmemeplex is 119898 The population iterative is iter Thelocal search iteration is mrun The maximum andminimum frog moving distances are 119863max and119863min respectively The crossover probability is pcThe adaptive mutation probability is pm

(2) Modified grouping operator the frogs are sorted indescending order according to their fitness during thepreprocessing of the grouping operator in SFLA In

Start

(3) Memetic evolution

(5) Global shuffling process

Yes

(2) Modified grouping operator

(1) Population initialization and parameter initialization

Convergence criteria are satisfied

Output the optimal solution

End

No

(4) Diversity control strategy

(local search based on swap sequence)

Figure 3 Memetic evolution procedure in SFLA

ASP a small fitness function value indicates good frogposition and good assembly sequence Therefore thegrouping operator should be modified The modifiedgrouping operator is as follows Suppose that the scaleof the frog population is 119873 which is then dividedinto 119898 memeplexes All frogs in the population arearranged in ascending order according to the fitnessfunction valueThefirst frog enters the firstmemeplexand the second frog enters the secondmemeplex andso on until the 119898 frog enters the 119898th memeplexThe (119898 + 1)th frog is then assigned back to thefirst memeplex and so on All individual frogs areassigned according to the aforementioned rule

(3) Memetic evolution (local search based on a swapsequence) in the improved SFLA memetic evolutionis modified and performed by using a local searchstrategy based on a swap sequence until the mrungeneration

(4) Optimal sampling different strategy the optimal sam-pling different strategy is included in the improvedSFLA to avoid homoplasy for each memeplex

(5) Global shuffling process the global shuffling processof the improved SFLA is similar to SFLA and updatesthe best position 119865

119892of the frog population

(6) The next step is determining whether the iterationshould be terminated according to the terminal con-dition of the algorithm If the terminal condition is

8 Science and Technology of Nuclear Installations

Table 1 Part assembly tool sets

Part number Part name Assembly tool1 Nut washer assembly 1 1198791

2 Nut washer assembly 2 1198791

3 Nut washer assembly 3 1198791

4 Nut washer assembly 4 1198791

5 Hydraulic cylinder 1198794

6 Pole 1 1198793

7 Pole 2 1198793

8 Pole 3 1198793

9 Pole 4 1198793

10 Strut 1 1198794

11 Nut 1 1198791

12 Nut 2 1198791

13 Nut 3 11987911198792

14 Bolt 1 11987911198793

15 Bolt 2 11987911198793

16 Pin 1 1198793

17 Nut 4 11987911198792

18 Pin 2 1198793

19 Central pin 1198793

20 Back plate 1198794

21 Strut 2 1198794

22 Nut washer assembly 5 1198791

23 Nut washer assembly 6 1198791

24 Nut washer assembly 7 1198791

25 Nut washer assembly 8 1198791

26 Axis 1198793

27 Hydraulic pressure scissors 1198794

28 Hydraulic pressure shear blades 1 1198795

29 Hydraulic pressure shear blades 2 1198795

satisfied then the iteration ends Otherwise return tostep (2)

6 Experiment and Analysis

The application program based on the improved SFLAis compiled under MATLAB environment The computerenvironment of the application program consists of a 20GHzCPU 2GBmemory andWindows 7 32-bit operating systemThe hydraulic pressure shear which contains 29 parts is usedfor the ASP experiment The exploded view of the hydraulicpressure shear is shown in Figure 4 The components of theassembly tool sets are listed in Table 1

61 ASP Experiment Based on SFLA After conducting anorthogonal experiment on the assembly of the hydraulicpressure shear the algorithm rapidly identifies an optimalassembly sequence when the weighting factors of the eval-uating indicator in the fitness function are 119888

119891= 4 119888

119904=

05 119888119905= 02 and 119888

119889= 03 If the memeplex has few local

search iterations then it also undergoes few evolution timeswhich reduces information exchange within the memeplex

If the memeplex has many local search iterations then itundergoes multiple local searches which increases algorithmsearch time and makes the best frog position of variousmemeplexes similar It also causes the algorithm to carry outGA several times and thus slows down the convergence rate ofthe algorithm If the maximum moving distance 119863max is toosmall then the global algorithm search capability is reducedIt causes the algorithm to easily fall into a local search If119863maxis too large then the algorithm is unable to find the globallyoptimal solutionAftermultiple comparison experiments thealgorithm optimization capability is observed to be optimalwhen the local search iteration of the memeplex is 10crossover probability is 08 adaptive mutation probabilityis 01 maximum moving distance is 8 minimum movingdistance is 1 and frog quantity in thememeplex ismaintainedat 30

The ASP experiment is conducted with population sizes60 120 180 and 240 given that each parameter value of theimproved SFLA and the weighting factor of the evaluatingindicator in the fitness function are the same The fitnessfunction value distributed the optimal assembly sequencefrom the results of the ASP experiment The analysis resultsare shown in Figure 5 in which the number of algorithmiterations is 600 and that of repeating operation times is 50A lot of experiments show that the fitness value of the globaloptimal assembly sequence 119865 is 21 As shown in Figure 5the distributed situation of the local optimal fitness value iswithin the following ranges 21 to 30 31 to 40 41 to 5051 to 60 and gt60 As shown in Figure 5 the distributedsituations of fitness value of local optimal assembly sequencediffer along with various population sizes When populationsize is 60 and time of experiment is 50 there is only onefitness value of local optimal assembly sequence in sections21 to 30 As population size increases the quantity offitness value of local optimal assembly sequence identifiedby the algorithm in this section gradually increases Whenpopulation size increases to 240 the quantity identifiedby the algorithm in this section is 20 As population sizeincrease the quantity of outstanding assembly sequenceswhose fitness value is smaller gradually increases As shownin Table 2 the increase in algorithm population size reducesalgorithm iteration efficiency and the operation time of thealgorithm is extended In this experiment the probabilityof the global optimal assembly sequence identified by thealgorithm is highest when population size is 240 and theaverage consuming time of a single experiment is in the rangeof acceptable with 515S

The mean and optimal average fitness of iteration in50 experiments when population size is 120 are shown inFigure 6 As algorithm iteration increases the optimal fitnessaverage value fluctuates However the optimal and averagevalues of the average fitness steadily decrease from the overalltendency which shows that the stability of the algorithm isgood and the algorithm convergence rate at the later periodis slow

The mean and optimal average fitness of 1 of the 50experiments in which the global optimal assembly sequenceis obtained when population size is 120 is shown in Figure 7In the algorithm implementation the mean fitness exhibits a

Science and Technology of Nuclear Installations 9

1

2

3 45

6

7

89

10

11

12

13

14

1516

1718

19

20

21

22

23

24

25

26

2728

29

Z

X

Y

Figure 4 Exploded view of the hydraulic pressure shear

huge fluctuation because the optimal sample is homoplasyAccording to the diversity control strategy the algorithmexecutes GA and increases average fitness The homoplasy ofthe algorithm improves after executing GA

62 Algorithm Comparison Experiment An algorithm com-parison experiment is conducted among improved SFLA(SFLA-GA) GA SFLA PSO and AMPSO [7] to verify theperformance of the improved SFLA for the ASP problemAMPSO ismodifiedmethod for ASP in RHM in our previouswork [7]

A hydraulic pressure shear is employed to carry out ASPwith the same programming and PC environments as thoseindicated in Section 61 Moreover the parameter 119891

119889= 21

and values of other related parameters in AMPSO are thesame as those presented in [11] Inertia weight in PSO is 06and the values of the other parameters are similar to thoseof AMPSO The experiment results with population sizes of60 and 240 are shown in Table 3 The algorithm convergencecurves of different population sizes are shown in Figures 8and 9 (ie variation of population average fitness values alongwith the iterations)

As shown in Table 3 the probability of a feasible assemblysequence identified by the improved SFLA is enhanced whenpopulation size increasesThe probability is higher than thosefor SFLA and GA but lower than those for AMPSO andPSO Under the same population size the improved SFLAexhibits a superior assembly sequence than those of GA andSFLA The value of the optimal assembly sequence fitnessfunction identified by the improved SFLA is less than those

identified by GA and SFLA When population size is 60the improved SFLA obtains an acceptable assembly sequenceand the fitness function value is 29 However GA and SFLAare unable to obtain an acceptable assembly sequence evenwhen the population size is 240 The fitness values of theoptimal assembly sequence are similar in PSO and AMPSOThe execution time of the improved SFLA is slightly less thanthat of GA and significantly less than those of AMPSO andPSO Consequently the efficiency of the proposed algorithmis acceptable Based on the local optimal fitness averagevalue of the algorithm the improved SFLA exhibits a higherconvergence rate than those of GA and SFLA and is nearthose of AMPSO and PSO The improved SFLA is evenbetter than PSO when population size is 240 Therefore theoptimization capability efficiency and convergence rate ofthe improved SFLA are better than those of GA whereasits optimization capability and convergence rate are betterthan those of SFLA The overall performance of the SFLA-GA for solving ASP problems proposed in this study issimilar to those of AMPSO and PSO As shown in Figures8 and 9 the stochastic initializing population qualities ofthe five algorithms are approximately similar Therefore thepreceding analysis is reliable

7 Conclusions

RHM is an important mean of ensuring the reliabilityof radioactive equipment and has a wide application inradioactive installations RHMP predetermine the mainte-nance procedures of radioactive equipment during the design

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

8 Science and Technology of Nuclear Installations

Table 1 Part assembly tool sets

Part number Part name Assembly tool1 Nut washer assembly 1 1198791

2 Nut washer assembly 2 1198791

3 Nut washer assembly 3 1198791

4 Nut washer assembly 4 1198791

5 Hydraulic cylinder 1198794

6 Pole 1 1198793

7 Pole 2 1198793

8 Pole 3 1198793

9 Pole 4 1198793

10 Strut 1 1198794

11 Nut 1 1198791

12 Nut 2 1198791

13 Nut 3 11987911198792

14 Bolt 1 11987911198793

15 Bolt 2 11987911198793

16 Pin 1 1198793

17 Nut 4 11987911198792

18 Pin 2 1198793

19 Central pin 1198793

20 Back plate 1198794

21 Strut 2 1198794

22 Nut washer assembly 5 1198791

23 Nut washer assembly 6 1198791

24 Nut washer assembly 7 1198791

25 Nut washer assembly 8 1198791

26 Axis 1198793

27 Hydraulic pressure scissors 1198794

28 Hydraulic pressure shear blades 1 1198795

29 Hydraulic pressure shear blades 2 1198795

satisfied then the iteration ends Otherwise return tostep (2)

6 Experiment and Analysis

The application program based on the improved SFLAis compiled under MATLAB environment The computerenvironment of the application program consists of a 20GHzCPU 2GBmemory andWindows 7 32-bit operating systemThe hydraulic pressure shear which contains 29 parts is usedfor the ASP experiment The exploded view of the hydraulicpressure shear is shown in Figure 4 The components of theassembly tool sets are listed in Table 1

61 ASP Experiment Based on SFLA After conducting anorthogonal experiment on the assembly of the hydraulicpressure shear the algorithm rapidly identifies an optimalassembly sequence when the weighting factors of the eval-uating indicator in the fitness function are 119888

119891= 4 119888

119904=

05 119888119905= 02 and 119888

119889= 03 If the memeplex has few local

search iterations then it also undergoes few evolution timeswhich reduces information exchange within the memeplex

If the memeplex has many local search iterations then itundergoes multiple local searches which increases algorithmsearch time and makes the best frog position of variousmemeplexes similar It also causes the algorithm to carry outGA several times and thus slows down the convergence rate ofthe algorithm If the maximum moving distance 119863max is toosmall then the global algorithm search capability is reducedIt causes the algorithm to easily fall into a local search If119863maxis too large then the algorithm is unable to find the globallyoptimal solutionAftermultiple comparison experiments thealgorithm optimization capability is observed to be optimalwhen the local search iteration of the memeplex is 10crossover probability is 08 adaptive mutation probabilityis 01 maximum moving distance is 8 minimum movingdistance is 1 and frog quantity in thememeplex ismaintainedat 30

The ASP experiment is conducted with population sizes60 120 180 and 240 given that each parameter value of theimproved SFLA and the weighting factor of the evaluatingindicator in the fitness function are the same The fitnessfunction value distributed the optimal assembly sequencefrom the results of the ASP experiment The analysis resultsare shown in Figure 5 in which the number of algorithmiterations is 600 and that of repeating operation times is 50A lot of experiments show that the fitness value of the globaloptimal assembly sequence 119865 is 21 As shown in Figure 5the distributed situation of the local optimal fitness value iswithin the following ranges 21 to 30 31 to 40 41 to 5051 to 60 and gt60 As shown in Figure 5 the distributedsituations of fitness value of local optimal assembly sequencediffer along with various population sizes When populationsize is 60 and time of experiment is 50 there is only onefitness value of local optimal assembly sequence in sections21 to 30 As population size increases the quantity offitness value of local optimal assembly sequence identifiedby the algorithm in this section gradually increases Whenpopulation size increases to 240 the quantity identifiedby the algorithm in this section is 20 As population sizeincrease the quantity of outstanding assembly sequenceswhose fitness value is smaller gradually increases As shownin Table 2 the increase in algorithm population size reducesalgorithm iteration efficiency and the operation time of thealgorithm is extended In this experiment the probabilityof the global optimal assembly sequence identified by thealgorithm is highest when population size is 240 and theaverage consuming time of a single experiment is in the rangeof acceptable with 515S

The mean and optimal average fitness of iteration in50 experiments when population size is 120 are shown inFigure 6 As algorithm iteration increases the optimal fitnessaverage value fluctuates However the optimal and averagevalues of the average fitness steadily decrease from the overalltendency which shows that the stability of the algorithm isgood and the algorithm convergence rate at the later periodis slow

The mean and optimal average fitness of 1 of the 50experiments in which the global optimal assembly sequenceis obtained when population size is 120 is shown in Figure 7In the algorithm implementation the mean fitness exhibits a

Science and Technology of Nuclear Installations 9

1

2

3 45

6

7

89

10

11

12

13

14

1516

1718

19

20

21

22

23

24

25

26

2728

29

Z

X

Y

Figure 4 Exploded view of the hydraulic pressure shear

huge fluctuation because the optimal sample is homoplasyAccording to the diversity control strategy the algorithmexecutes GA and increases average fitness The homoplasy ofthe algorithm improves after executing GA

62 Algorithm Comparison Experiment An algorithm com-parison experiment is conducted among improved SFLA(SFLA-GA) GA SFLA PSO and AMPSO [7] to verify theperformance of the improved SFLA for the ASP problemAMPSO ismodifiedmethod for ASP in RHM in our previouswork [7]

A hydraulic pressure shear is employed to carry out ASPwith the same programming and PC environments as thoseindicated in Section 61 Moreover the parameter 119891

119889= 21

and values of other related parameters in AMPSO are thesame as those presented in [11] Inertia weight in PSO is 06and the values of the other parameters are similar to thoseof AMPSO The experiment results with population sizes of60 and 240 are shown in Table 3 The algorithm convergencecurves of different population sizes are shown in Figures 8and 9 (ie variation of population average fitness values alongwith the iterations)

As shown in Table 3 the probability of a feasible assemblysequence identified by the improved SFLA is enhanced whenpopulation size increasesThe probability is higher than thosefor SFLA and GA but lower than those for AMPSO andPSO Under the same population size the improved SFLAexhibits a superior assembly sequence than those of GA andSFLA The value of the optimal assembly sequence fitnessfunction identified by the improved SFLA is less than those

identified by GA and SFLA When population size is 60the improved SFLA obtains an acceptable assembly sequenceand the fitness function value is 29 However GA and SFLAare unable to obtain an acceptable assembly sequence evenwhen the population size is 240 The fitness values of theoptimal assembly sequence are similar in PSO and AMPSOThe execution time of the improved SFLA is slightly less thanthat of GA and significantly less than those of AMPSO andPSO Consequently the efficiency of the proposed algorithmis acceptable Based on the local optimal fitness averagevalue of the algorithm the improved SFLA exhibits a higherconvergence rate than those of GA and SFLA and is nearthose of AMPSO and PSO The improved SFLA is evenbetter than PSO when population size is 240 Therefore theoptimization capability efficiency and convergence rate ofthe improved SFLA are better than those of GA whereasits optimization capability and convergence rate are betterthan those of SFLA The overall performance of the SFLA-GA for solving ASP problems proposed in this study issimilar to those of AMPSO and PSO As shown in Figures8 and 9 the stochastic initializing population qualities ofthe five algorithms are approximately similar Therefore thepreceding analysis is reliable

7 Conclusions

RHM is an important mean of ensuring the reliabilityof radioactive equipment and has a wide application inradioactive installations RHMP predetermine the mainte-nance procedures of radioactive equipment during the design

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Science and Technology of Nuclear Installations 9

1

2

3 45

6

7

89

10

11

12

13

14

1516

1718

19

20

21

22

23

24

25

26

2728

29

Z

X

Y

Figure 4 Exploded view of the hydraulic pressure shear

huge fluctuation because the optimal sample is homoplasyAccording to the diversity control strategy the algorithmexecutes GA and increases average fitness The homoplasy ofthe algorithm improves after executing GA

62 Algorithm Comparison Experiment An algorithm com-parison experiment is conducted among improved SFLA(SFLA-GA) GA SFLA PSO and AMPSO [7] to verify theperformance of the improved SFLA for the ASP problemAMPSO ismodifiedmethod for ASP in RHM in our previouswork [7]

A hydraulic pressure shear is employed to carry out ASPwith the same programming and PC environments as thoseindicated in Section 61 Moreover the parameter 119891

119889= 21

and values of other related parameters in AMPSO are thesame as those presented in [11] Inertia weight in PSO is 06and the values of the other parameters are similar to thoseof AMPSO The experiment results with population sizes of60 and 240 are shown in Table 3 The algorithm convergencecurves of different population sizes are shown in Figures 8and 9 (ie variation of population average fitness values alongwith the iterations)

As shown in Table 3 the probability of a feasible assemblysequence identified by the improved SFLA is enhanced whenpopulation size increasesThe probability is higher than thosefor SFLA and GA but lower than those for AMPSO andPSO Under the same population size the improved SFLAexhibits a superior assembly sequence than those of GA andSFLA The value of the optimal assembly sequence fitnessfunction identified by the improved SFLA is less than those

identified by GA and SFLA When population size is 60the improved SFLA obtains an acceptable assembly sequenceand the fitness function value is 29 However GA and SFLAare unable to obtain an acceptable assembly sequence evenwhen the population size is 240 The fitness values of theoptimal assembly sequence are similar in PSO and AMPSOThe execution time of the improved SFLA is slightly less thanthat of GA and significantly less than those of AMPSO andPSO Consequently the efficiency of the proposed algorithmis acceptable Based on the local optimal fitness averagevalue of the algorithm the improved SFLA exhibits a higherconvergence rate than those of GA and SFLA and is nearthose of AMPSO and PSO The improved SFLA is evenbetter than PSO when population size is 240 Therefore theoptimization capability efficiency and convergence rate ofthe improved SFLA are better than those of GA whereasits optimization capability and convergence rate are betterthan those of SFLA The overall performance of the SFLA-GA for solving ASP problems proposed in this study issimilar to those of AMPSO and PSO As shown in Figures8 and 9 the stochastic initializing population qualities ofthe five algorithms are approximately similar Therefore thepreceding analysis is reliable

7 Conclusions

RHM is an important mean of ensuring the reliabilityof radioactive equipment and has a wide application inradioactive installations RHMP predetermine the mainte-nance procedures of radioactive equipment during the design

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

10 Science and Technology of Nuclear Installations

Table 2 Execution of the 50 optimal ASP comparison results under different population sizes

Population size 60 120 180 240S D T S D T S D T S D T

Assembly sequencerelated information

20 minus119883 1198794 5 minus119884 1198794 20 minus119884 1198794 20 minus119883 1198794

21 minus119883 1198794 20 minus119884 1198794 10 minus119884 1198794 10 minus119883 1198794

10 minus119883 1198794 10 minus119884 1198794 21 minus119884 1198794 5 minus119883 1198794

5 minus119883 1198794 21 minus119884 1198794 26 minus119884 1198793 21 minus119884 1198794

26 minus119884 1198793 7 minus119884 1198793 27 minus119884 1198794 8 minus119884 1198793

27 minus119884 1198794 6 minus119884 1198793 29 minus119884 1198795 26 minus119884 1198793

29 minus119884 1198795 9 minus119884 1198793 28 minus119884 1198795 6 minus119884 1198793

28 minus119884 1198795 26 minus119884 1198793 5 minus119885 1198794 9 minus119884 1198793

23 minus119885 1198791 8 minus119884 1198793 19 minus119885 1198793 7 minus119884 1198793

15 minus119885 1198791 22 minus119884 1198791 16 minus119885 1198793 22 minus119884 1198791

25 minus119885 1198791 24 minus119884 1198791 18 minus119885 1198793 25 minus119884 1198791

22 minus119885 1198791 23 minus119884 1198791 23 minus119885 1198791 24 minus119884 1198791

24 minus119885 1198791 25 minus119884 1198791 15 minus119885 1198791 23 minus119884 1198791

19 minus119885 1198793 27 minus119884 1198794 25 minus119885 1198791 27 minus119884 1198794

9 +119884 1198793 28 minus119884 1198795 22 minus119885 1198791 28 minus119884 1198795

6 +119884 1198793 29 minus119884 1198795 17 minus119885 1198791 29 minus119884 1198795

8 +119884 1198793 16 minus119885 1198793 14 minus119885 1198791 14 minus119885 1198793

7 +119884 1198793 18 minus119885 1198793 24 minus119885 1198791 18 minus119885 1198793

1 +119884 1198791 14 minus119885 1198793 6 +119884 1198793 19 minus119885 1198793

4 +119884 1198791 19 minus119885 1198793 9 +119884 1198793 16 minus119885 1198793

3 +119884 1198791 17 minus119885 1198791 7 +119884 1198793 17 minus119885 1198791

2 +119884 1198791 15 minus119885 1198791 8 +119884 1198793 15 minus119885 1198791

12 +119885 1198791 12 +119885 1198791 1 +119884 1198791 11 +119885 1198791

13 +119885 1198791 11 +119885 1198791 3 +119884 1198791 13 +119885 1198791

11 +119885 1198791 13 +119885 1198791 2 +119884 1198791 12 +119885 1198791

17 minus119885 1198791 3 +119884 1198791 4 +119884 1198791 2 +119884 1198791

16 minus119885 1198793 1 +119884 1198791 12 +119885 1198791 4 +119884 1198791

14 minus119885 1198793 4 +119884 1198791 11 +119885 1198791 3 +119884 1198791

18 minus119885 1198793 2 +119884 1198791 13 +119885 1198791 1 +119884 1198791

Changing times ofassemble direction 5 3 3 4

Changing times ofassemble tool 7 6 8 6

Unstable operationtimes 0 2 0 0

Single executiontimes 135 256 387 515

Fitness 29 21 25 24Note 119878 represents the assembly sequence119863 represents the assembly direction and 119879 represents the assembly tool

of radioactive installation As a part of RHMP ASP isintroduced in this study Evolution algorithm is a useful toolfor ASP which is considered as a combinatorial optimizationproblem The contribution of this study is to develop anadvanced evolution algorithm named improved SFLA forASP

SFLA is an evolution algorithm that is used to calculatethe global optima of several combinatorial problems andhas been found to be effective in searching for global

solutions There are mainly two improvement strategies thatwere employed in our improved algorithm (1) each SFLAoperation was redefined with respect to the discretenesscharacteristic of ASP (2) a diversity control strategy basedon GA was introduced to avoid homoplasy for each meme-plex The experiments proved that the global optimizationcapability and convergence rate of the improved SFLA arebetter than those of SFLA and GA and similar to thoseof AMPSO and PSO Moreover the algorithm operation

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Science and Technology of Nuclear Installations 11

Tim

es

0

10

20

Fitness2 4 6

(a) Population size is 60

Tim

es

0

10

20

Fitness2 4 6

(b) Population size is 120

Fitness2

Tim

es

0

10

20

4 6

(c) Population size is 180

Tim

es

0

10

20

Fitness2 4 6

(d) Population size is 240

Figure 5 Distributed situations of each local optimal fitness

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean average fitnessOptimal average fitness

Figure 6 Mean and optimal average fitness of iteration in 50experiments

efficiency should be better than GA to achieve an enhancedassembly sequence result Experiment results showed that

0 100 200 300 400 500 6000

102030405060

Generation

Fitn

ess

Mean fitnessOptimal fitness

Figure 7 Mean and optimal fitness

the proposed algorithm is an advanced evolution algorithmand exhibits outstanding performance in solving the ASP

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

12 Science and Technology of Nuclear Installations

Table 3 Comparison test results of the algorithm

Parameter GA SFLA AMPSO PSO SFLAndashGAPopulation size 60 240 60 240 60 240 60 240 60 240Iteration times 600 600 600 600 600 600 600 600 600 600Execution times 50 50 50 50 50 50 50 50 50 50Feasible sequence number 3 1 0 1 33 40 29 40 19 37Running times 7456 29586 6920 27762 8578 34822 8484 34178 6761 25763Optimal assembly sequencefitness value 70 64 68 65 24 24 350 23 29 24

Local optimal-fitnessaverage value 1008 817 1277 998 527 379 516 413 577 374

0 100 200 300 400 500 6000

10

20

30

40

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 8 Convergence curve (population size = 60)

0 100 200 300 400 500 60005

1015202530

Generation

Fitn

ess

GAPSOSFLA

AMPSOSFLA-GA

Figure 9 Convergence curve (population size = 240)

problem The application of the proposed algorithm shouldincrease the level of ASP in a radioactive environmentHowever the experiments also proved that the convergencerate of the improved SFLA at the later period is slowFurther works should be conducted to improve the proposedalgorithm and enhance its convergence rate

Notations

AMPSO Adaptive mutation particle swarmoptimization

AP Assemble sequence

ASP Assembly sequence planning119888119894119895 Connection type between part 119875

119894and part

119875119895

119863119862(119875119894) Feasible assembly direction set of part 119875

119894

in an assemble sequence119889119896 The 119896th direction in the set

119889(119896) = +119909 +119910 +119911 minus119909 minus119910 minus119911

119863max Maximum distance that the frog ispermitted to move

119865119887 The best positions of each frog in each

memeplex119865119908 The worst positions of each frog in each

memeplexGA Genetic algorithm119868119894119895119889119896

Whether part 119875119894interfere with part 119875

119895

when moving along 119889119896direction

ITER International thermonuclear experimentalreactor

119899119889 Changing times of assembly direction119899119891 Total times of assembly interference in an

assembly sequence119899119904 Times of the assembly sequence stable

operation119899119905 Changing times of assembly tool119875119894 The 119894th part in an assemble sequence

PSO Particle swarm optimizationRHM Remote handling maintenanceRHMP Remote handling maintenance planningSFLA Shuffled frog leaping algorithmSFLA-GA Improved shuffled frog leaping algorithm

proposed by this paper119904119894119895 Support type between part 119875

119894and part 119875

119895

119878119896(119875119894) Sum of the interference values

119879119888(119875119894) The assembly tool of part 119875

119894

V119900(120574 120596) Swap factor to swap the positions of 120574 and120596 to form a new assembly sequence

V119900119904 Swap sequence which is swap factors set

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Science and Technology of Nuclear Installations 13

Acknowledgments

The study was supported by the National Natural ScienceFoundation of China (Grant no 71201026) the Project ofthe Department of Education of Guangdong Province (no2013KJCX0179 no 2014KTSCX184 and no 2014KGJHZ014)the Development Program for Excellent Young Teachers inHigher Education Institutions of Guangdong Province (noYq2013156) the China Spallation Neutron Source Electrome-chanical Technology RampD Joint Laboratory Foundation (noZD120512) the Dongguan Social Science and TechnologyDevelopment Project (no 2013108101011) and the DongguanUniversities and Scientific Research Institutions Science andTechnology Project (no 2014106101007)

References

[1] J L Seminara and S O Parsons ldquoNuclear power plant main-tainabilityrdquo Applied Ergonomics vol 13 no 3 pp 177ndash189 1982

[2] W B Scott W I Enderlin A D Chockie and K A BjorkeloldquoGood practices for effective maintenance to manage aging ofnuclear power plantsrdquoNuclear Engineering and Design vol 134no 2-3 pp 257ndash265 1992

[3] S Martorell V Serradell and G Verdu ldquoSafety-related equip-ment prioritization for reliability centered maintenance pur-poses based on a plant specific level 1 PSArdquo Reliability Engineer-ing and System Safety vol 52 no 1 pp 35ndash44 1996

[4] H Kinoshita M Teshigawara M Ito et al ldquoRemote handlingdevices in MLFrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 600 no 1 pp 78ndash80 2009

[5] EUROfusion httpswwweuro-fusionorgfusionjet-remote-handlingwhat-is-remote-handling

[6] I Ribeiro C Damiani A Tesini S Kakudate M Siuko andC Neri ldquoThe remote handling systems for ITERrdquo FusionEngineering and Design vol 86 no 9-10 pp 471ndash477 2011

[7] J-WGuo SWangH-B Chen Z-Z Sun andZ-C Zhang ldquoAnintelligent approach for remote handlingmaintenance sequenceplanning in radiation environmentrdquo Journal of Chemical andPharmaceutical Research vol 6 no 4 pp 543ndash552 2014

[8] L Wang S Keshavarzmanesh H-Y Feng and R-O BuchalldquoAssembly process planning and its future in collaborativemanufacturing a reviewrdquoThe International Journal of AdvancedManufacturing Technology vol 41 no 1-2 pp 132ndash144 2009

[9] R H Wilson and J-C Latombe ldquoGeometric reasoning aboutmechanical assemblyrdquo Artificial Intelligence vol 71 no 2 pp371ndash396 1994

[10] L Kavraki J-C Latombe andRHWilson ldquoOn the complexityof assembly partitioningrdquo Information Processing Letters vol48 no 5 pp 229ndash235 1993

[11] P Jimenez ldquoSurvey on assembly sequencing a combinatorialand geometrical perspectiverdquo Journal of Intelligent Manufactur-ing vol 24 no 2 pp 235ndash250 2013

[12] MM Eusuff and K E Lansey ldquoOptimization of water distribu-tion network design using the shuffled frog leaping algorithmrdquoJournal of Water Resources Planning and Management vol 129no 3 pp 210ndash225 2003

[13] M Eusuff K Lansey and F Pasha ldquoShuffled frog-leapingalgorithm a memetic meta-heuristic for discrete optimizationrdquoEngineering Optimization vol 38 no 2 pp 129ndash154 2006

[14] Japan Atomic Energy Agency Technical Design Report of Spal-lation Neutron Source Facility in J-PARC Japan Atomic EnergyAgency Ibaraki-ken Japan 2012

[15] Z Zhou D Yao and P Zi ldquoThe research activities on remotehandling system for CFETRrdquo Journal of Fusion Energy vol 34no 2 pp 232ndash237 2015

[16] K Kershaw B Feral J-L Grenard et al ldquoRemote inspectionmeasurement and handling for maintenance and operation atCERNrdquo International Journal of Advanced Robotic Systems vol10 no 382 pp 1ndash11 2013

[17] N Takeda K Akou S Kakudate et al ldquoDevelopment ofdivertor cassette transporters for ITERrdquo in Proceedings of the17th IEEENPSS Symposium on Fusion Engineering vol 2 pp925ndash928 October 1997

[18] P Desbats F Geffard G Piolain and A Coudray ldquoForce-feedback teleoperation of an industrial robot in a nuclear spentfuel reprocessing plantrdquo Industrial Robot vol 33 no 3 pp 178ndash186 2006

[19] F Geffard P Garrec G Piolain et al ldquoTAO2000 V2 computer-assisted force feedback telemanipulators used as maintenanceand production tools at theAREVANC-LaHague fuel recyclingplantrdquo Journal of Field Robotics vol 29 no 1 pp 161ndash174 2012

[20] S Sanders ldquoRemote operations for fusion using teleoperationrdquoIndustrial Robot vol 33 no 3 pp 174ndash177 2006

[21] A Vale D Fonte F Valente and I Ribeiro ldquoTrajectoryoptimization for autonomous mobile robots in ITERrdquo Roboticsand Autonomous Systems vol 62 no 6 pp 871ndash888 2014

[22] S Terada H Kobayashi H Sengoku et al ldquoDesign anddevelopment of awork robot to placeATLASSCTmodules ontobarrel cylindersrdquo Nuclear Instruments and Methods in PhysicsResearch Section A Accelerators Spectrometers Detectors andAssociated Equipment vol 541 no 1-2 pp 144ndash149 2005

[23] H J Lee J K Lee B S Park and J S Yoon ldquoBridge transportedservo manipulator system for remote handling tasks under aradiation environmentrdquo Industrial Robot vol 36 no 2 pp 165ndash175 2009

[24] J K Lee B S Park K Kim and H D Kim ldquoDesign andfabrication of a servo-manipulator for use in the PRIDE facilityrdquoin Proceedings of IEEE International Symposium on Assemblyand Manufacturing (ISAM rsquo09) pp 417ndash421 November 2009

[25] B Elzendoorn M-D Baar R Chavan et al ldquoAnalysis of theITER ECH Upper Port Launcher remote maintenance usingvirtual realityrdquo Fusion Engineering and Design vol 84 no 2ndash6pp 733ndash735 2009

[26] N Takeda S Kakudate M Nakahira K Shibanuma and ATesini ldquoDevelopment of a virtual reality simulator for theITER blanket remote handling systemrdquo Fusion Engineering andDesign vol 83 no 10ndash12 pp 1837ndash1840 2008

[27] C J M Heemskerk M R De Baar H Boessenkool et alldquoExtending Virtual Reality simulation of ITER maintenanceoperations with dynamic effectsrdquo Fusion Engineering andDesign vol 86 no 9-11 pp 2082ndash2086 2011

[28] J Geng D Zhou C Lv and Z Wang ldquoA modeling approachfor maintenance safety evaluation in a virtual maintenanceenvironmentrdquo Computer Aided Design vol 45 no 5 pp 937ndash949 2013

[29] S Esque J Mattila M Siuko et al ldquoThe use of digital mock-ups on the development of the Divertor Test Platform 2rdquo FusionEngineering and Design vol 84 no 2ndash6 pp 752ndash756 2009

[30] R Shuff D Locke and D A Roulet ldquoITER Hot Cell processoperability analysis using discrete event simulation toolsrdquo

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

14 Science and Technology of Nuclear Installations

Fusion Engineering and Design vol 86 no 9ndash11 pp 1950ndash19532011

[31] E Robbins S Sanders A Williams and P Allan ldquoThe use ofvirtual reality and intelligent database systems for procedureplanning visualisation and real-time component tracking inremote handling operationsrdquo Fusion Engineering and Designvol 84 no 7ndash11 pp 1628ndash1632 2009

[32] H-S Park S-K Kim K-W Lee C-H Jung J-H Park andS-I Jin ldquoThe application of visualization and simulation in adismantling processrdquo Journal of Nuclear Science and Technologyvol 44 no 4 pp 649ndash656 2007

[33] H S Park CHChoi SHKim B S Park KHKim andHDKim ldquoDeployment analysis and remote accessibility verificationfor a maintenance task in a PRIDE digital mock-uprdquo Annals ofNuclear Energy vol 38 no 4 pp 767ndash774 2011

[34] A R Selva R C Castro and V G F Frıas ldquoDesign ofdisassembly sequences using search strategies Application ofIDA in state diagramsrdquo International Journal of ProductionResearch vol 49 no 11 pp 3395ndash3403 2011

[35] Y Xing G Chen X Lai S Jin and J Zhou ldquoAssembly sequenceplanning of automobile body components based on liaisongraphrdquo Assembly Automation vol 27 no 2 pp 157ndash164 2007

[36] J Ko E Nazarian H Wang and J Abell ldquoAn assemblydecomposition model for subassembly planning consideringimperfect inspection to reduce assembly defect ratesrdquo Journalof Manufacturing Systems vol 32 no 3 pp 412ndash416 2013

[37] G Dini and M Santochi ldquoAutomated sequencing and sub-assembly detection in assembly planningrdquo CIRP AnnalsmdashManufacturing Technology vol 41 no 1 pp 1ndash4 1992

[38] C Morato K N Kaipa and S K Gupta ldquoImproving assemblyprecedence constraint generation by utilizing motion planningand part interaction clustersrdquo Computer-Aided Design vol 45no 11 pp 1349ndash1364 2013

[39] T G Chen and R B Xiao ldquoEnhancing artificial bee colonyalgorithm with self-adaptive searching strategy and artificialimmune network operators for global optimizationrdquoThe Scien-tific World Journal vol 2014 Article ID 438260 12 pages 2014

[40] H Xu L Xin H Wang and N Yang ldquoDistribution networkplanning based on improved genetic algorithmrdquo GuangdongElectric Power vol 23 no 6 pp 6ndash9 2010

[41] T-G Chen andC-H Ju ldquoAnovel artificial bee colony algorithmfor solving the supply chain network design under disruptionscenariosrdquo International Journal of Computer Applications inTechnology vol 47 no 2-3 pp 289ndash296 2013

[42] C-H Cheng S C Ho and C-L Kwan ldquoThe use of meta-heuristics for airport gate assignmentrdquo Expert Systems withApplications vol 39 no 16 pp 12430ndash12437 2012

[43] S Lorpunmance and A Sap ldquoAn ant colony optimization fordynamic job scheduling in grid environmentrdquo InternationalJournal of Computer and Information Science and Engineeringvol 1 pp 207ndash214 2007

[44] P De Lit P Latinne B Rekiek and A Delchambre ldquoAssemblyplanning with an ordering genetic algorithmrdquo InternationalJournal of Production Research vol 39 no 16 pp 3623ndash36402001

[45] H-E Tseng W-P Wang and H-Y Shih ldquoUsing memeticalgorithms with guided local search to solve assembly sequenceplanningrdquo Expert Systems with Applications vol 33 no 2 pp451ndash467 2007

[46] H S Wang Z H Che and C J Chiang ldquoA hybrid geneticalgorithm for multi-objective product plan selection problem

with ASP and ALBrdquo Expert Systems with Applications vol 39no 5 pp 5440ndash5450 2012

[47] J-F Wang J-H Liu and Y-F Zhong ldquoA novel ant colonyalgorithm for assembly sequence planningrdquo The InternationalJournal of Advanced Manufacturing Technology vol 25 no 11-12 pp 1137ndash1143 2005

[48] H Shan S-H Zhou and Z-H Sun ldquoResearch on assem-bly sequence planning based on genetic simulated annealingalgorithm and ant colony optimization algorithmrdquo AssemblyAutomation vol 29 no 3 pp 249ndash256 2009

[49] Y Wang and J H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[50] M Li B Wu Y Hu C Jin and T Shi ldquoA hybrid assemblysequence planning approach based on discrete particle swarmoptimization and evolutionary direction operationrdquo The Inter-national Journal of AdvancedManufacturing Technology vol 68no 1ndash4 pp 617ndash630 2013

[51] Y Wang and J-H Liu ldquoChaotic particle swarm optimizationfor assembly sequence planningrdquo Robotics and Computer-Integrated Manufacturing vol 26 no 2 pp 212ndash222 2010

[52] M Li B Wu P Yi C Jin Y Hu and T Shi ldquoAn improveddiscrete particle swarm optimization algorithm for high-speedtrains assembly sequence planningrdquo Assembly Automation vol33 no 4 pp 360ndash373 2013

[53] G Percoco and M Diella ldquoPreliminary evaluation of artificialbee colony algorithm when applied to multi objective partialdisassembly planningrdquo Research Journal of Applied SciencesEngineering and Technology vol 6 no 17 pp 3234ndash3243 2013

[54] A Rahimi-Vahed and A H Mirzaei ldquoA hybrid multi-objectiveshuffled frog-leaping algorithm for a mixed-model assemblyline sequencing problemrdquo Computers amp Industrial Engineeringvol 53 no 4 pp 642ndash666 2007

[55] A Rahimi-Vahed M Dangchi H Rafiei and E Salimi ldquoAnovel hybrid multi-objective shuffled frog-leaping algorithmfor a bi-criteria permutation flow shop scheduling problemrdquoInternational Journal of Advanced Manufacturing Technologyvol 41 no 11-12 pp 1227ndash1239 2009

[56] X Li J-P Luo M-R Chen and N Wang ldquoAn improvedshuffled frog-leaping algorithm with extremal optimisation forcontinuous optimisationrdquo Information Sciences vol 192 no 1pp 143ndash151 2012

[57] C Fang and L Wang ldquoAn effective shuffled frog-leaping algo-rithm for resource-constrained project scheduling problemrdquoComputers amp Operations Research vol 39 no 5 pp 890ndash9012012

[58] J-Q Li Q Pan and S-X Xie ldquoAn effective shuffled frog-leapingalgorithm for multi-objective flexible job shop schedulingproblemsrdquo Applied Mathematics and Computation vol 218 no18 pp 9353ndash9371 2012

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

TribologyAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

FuelsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal ofPetroleum Engineering

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Industrial EngineeringJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Power ElectronicsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Advances in

CombustionJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Renewable Energy

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

StructuresJournal of

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear InstallationsScience and Technology of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Solar EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Wind EnergyJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Nuclear EnergyInternational Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

High Energy PhysicsAdvances in

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014