reliability-based structural optimization for positioning of marine vessels

31
Cesos-Workshop-March- 2006 1 RELIABILITY-BASED STRUCTURAL OPTIMIZATION FOR POSITIONING OF MARINE VESSELS B. J. Leira, NTNU, Trondheim, Norway P. I. B. Berntsen, NTNU, Trondheim, Norway O. M. Aamo, NTNU, Trondheim, Norway

Upload: overton

Post on 22-Jan-2016

41 views

Category:

Documents


1 download

DESCRIPTION

RELIABILITY-BASED STRUCTURAL OPTIMIZATION FOR POSITIONING OF MARINE VESSELS. B. J. Leira, NTNU, Trondheim, Norway P. I. B. Berntsen, NTNU, Trondheim, Norway O. M. Aamo, NTNU, Trondheim, Norway. Objective. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 1

RELIABILITY-BASED STRUCTURAL OPTIMIZATION FOR

POSITIONING OF MARINE VESSELS

B J Leira NTNU Trondheim Norway

P I B Berntsen NTNU Trondheim Norway

O M Aamo NTNU Trondheim Norway

Cesos-Workshop-March-2006 2

Objective

bull To investigate the possibility of implementing structural response and design criteria into the Dynamic Positioning control loop

bull Use a simplified quasistatic response model to derive optimal reliability levels for PID and LQG control schemes in conjunction with two different types of loss functions

bull Implement a control algorithm that is capable of achieving a given target reliability level for a realistic and fully dynamic system

Cesos-Workshop-March-2006 3

Control of low-frequency response level

Response

time

Cesos-Workshop-March-2006 4

Possible strategies for control algorithm based on

reliability indices

1 Monitoring of reliability indices

2 Weight factors based on reliability indices

3 Derivation of optimal control criteria based on reliability indices

Cesos-Workshop-March-2006 5

Principle Measure of structural safety is the reliability index β which is related to the failure probability by β = Φ-1(pf)

bull No activation β gt βThreshold

bull Alert interval increasing activation

βThreshold gt β gt βCritical

bull Full Activation β lt βCritical

Cesos-Workshop-March-2006 6

Definition of delta index

(ie due to waves)

is the mean breaking strength of the line

is the standard deviation of the breaking strength

Cesos-Workshop-March-2006 7

Computation of reliability index

bull Failure probability (pf) is probability that the extreme dynamic response will exceed critical level within a given reference duration

bull Failure probability is estimated for a stationary reference time interval of eg 20 minutes by application of a Gumbel distribution

bull Simplified relationship between delta-index and failure probability is expressed as

pf = (-δ)

Cesos-Workshop-March-2006 8

Simplified quasistatic loadresponse model is applied

for initial rdquooptimization studyrdquo

kTot∙r = FE - FT

where is total linearized stiffness of mooring lines FE is external (low-frequency) excitation and FT is thruster force

Conversely r = (FE ndash FT)kTot

Cesos-Workshop-March-2006 9

Two types of loss functions are considered

bull Typical LQG type of loss function

L( r ) = KTFT2 + KFr2

(r is response FT is thruster force KT and KF are constants)

bull Loss function based on failure probability

L( r ) = KT FT2 + KPФ(-δ)

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 2: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 2

Objective

bull To investigate the possibility of implementing structural response and design criteria into the Dynamic Positioning control loop

bull Use a simplified quasistatic response model to derive optimal reliability levels for PID and LQG control schemes in conjunction with two different types of loss functions

bull Implement a control algorithm that is capable of achieving a given target reliability level for a realistic and fully dynamic system

Cesos-Workshop-March-2006 3

Control of low-frequency response level

Response

time

Cesos-Workshop-March-2006 4

Possible strategies for control algorithm based on

reliability indices

1 Monitoring of reliability indices

2 Weight factors based on reliability indices

3 Derivation of optimal control criteria based on reliability indices

Cesos-Workshop-March-2006 5

Principle Measure of structural safety is the reliability index β which is related to the failure probability by β = Φ-1(pf)

bull No activation β gt βThreshold

bull Alert interval increasing activation

βThreshold gt β gt βCritical

bull Full Activation β lt βCritical

Cesos-Workshop-March-2006 6

Definition of delta index

(ie due to waves)

is the mean breaking strength of the line

is the standard deviation of the breaking strength

Cesos-Workshop-March-2006 7

Computation of reliability index

bull Failure probability (pf) is probability that the extreme dynamic response will exceed critical level within a given reference duration

bull Failure probability is estimated for a stationary reference time interval of eg 20 minutes by application of a Gumbel distribution

bull Simplified relationship between delta-index and failure probability is expressed as

pf = (-δ)

Cesos-Workshop-March-2006 8

Simplified quasistatic loadresponse model is applied

for initial rdquooptimization studyrdquo

kTot∙r = FE - FT

where is total linearized stiffness of mooring lines FE is external (low-frequency) excitation and FT is thruster force

Conversely r = (FE ndash FT)kTot

Cesos-Workshop-March-2006 9

Two types of loss functions are considered

bull Typical LQG type of loss function

L( r ) = KTFT2 + KFr2

(r is response FT is thruster force KT and KF are constants)

bull Loss function based on failure probability

L( r ) = KT FT2 + KPФ(-δ)

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 3: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 3

Control of low-frequency response level

Response

time

Cesos-Workshop-March-2006 4

Possible strategies for control algorithm based on

reliability indices

1 Monitoring of reliability indices

2 Weight factors based on reliability indices

3 Derivation of optimal control criteria based on reliability indices

Cesos-Workshop-March-2006 5

Principle Measure of structural safety is the reliability index β which is related to the failure probability by β = Φ-1(pf)

bull No activation β gt βThreshold

bull Alert interval increasing activation

βThreshold gt β gt βCritical

bull Full Activation β lt βCritical

Cesos-Workshop-March-2006 6

Definition of delta index

(ie due to waves)

is the mean breaking strength of the line

is the standard deviation of the breaking strength

Cesos-Workshop-March-2006 7

Computation of reliability index

bull Failure probability (pf) is probability that the extreme dynamic response will exceed critical level within a given reference duration

bull Failure probability is estimated for a stationary reference time interval of eg 20 minutes by application of a Gumbel distribution

bull Simplified relationship between delta-index and failure probability is expressed as

pf = (-δ)

Cesos-Workshop-March-2006 8

Simplified quasistatic loadresponse model is applied

for initial rdquooptimization studyrdquo

kTot∙r = FE - FT

where is total linearized stiffness of mooring lines FE is external (low-frequency) excitation and FT is thruster force

Conversely r = (FE ndash FT)kTot

Cesos-Workshop-March-2006 9

Two types of loss functions are considered

bull Typical LQG type of loss function

L( r ) = KTFT2 + KFr2

(r is response FT is thruster force KT and KF are constants)

bull Loss function based on failure probability

L( r ) = KT FT2 + KPФ(-δ)

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 4: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 4

Possible strategies for control algorithm based on

reliability indices

1 Monitoring of reliability indices

2 Weight factors based on reliability indices

3 Derivation of optimal control criteria based on reliability indices

Cesos-Workshop-March-2006 5

Principle Measure of structural safety is the reliability index β which is related to the failure probability by β = Φ-1(pf)

bull No activation β gt βThreshold

bull Alert interval increasing activation

βThreshold gt β gt βCritical

bull Full Activation β lt βCritical

Cesos-Workshop-March-2006 6

Definition of delta index

(ie due to waves)

is the mean breaking strength of the line

is the standard deviation of the breaking strength

Cesos-Workshop-March-2006 7

Computation of reliability index

bull Failure probability (pf) is probability that the extreme dynamic response will exceed critical level within a given reference duration

bull Failure probability is estimated for a stationary reference time interval of eg 20 minutes by application of a Gumbel distribution

bull Simplified relationship between delta-index and failure probability is expressed as

pf = (-δ)

Cesos-Workshop-March-2006 8

Simplified quasistatic loadresponse model is applied

for initial rdquooptimization studyrdquo

kTot∙r = FE - FT

where is total linearized stiffness of mooring lines FE is external (low-frequency) excitation and FT is thruster force

Conversely r = (FE ndash FT)kTot

Cesos-Workshop-March-2006 9

Two types of loss functions are considered

bull Typical LQG type of loss function

L( r ) = KTFT2 + KFr2

(r is response FT is thruster force KT and KF are constants)

bull Loss function based on failure probability

L( r ) = KT FT2 + KPФ(-δ)

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 5: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 5

Principle Measure of structural safety is the reliability index β which is related to the failure probability by β = Φ-1(pf)

bull No activation β gt βThreshold

bull Alert interval increasing activation

βThreshold gt β gt βCritical

bull Full Activation β lt βCritical

Cesos-Workshop-March-2006 6

Definition of delta index

(ie due to waves)

is the mean breaking strength of the line

is the standard deviation of the breaking strength

Cesos-Workshop-March-2006 7

Computation of reliability index

bull Failure probability (pf) is probability that the extreme dynamic response will exceed critical level within a given reference duration

bull Failure probability is estimated for a stationary reference time interval of eg 20 minutes by application of a Gumbel distribution

bull Simplified relationship between delta-index and failure probability is expressed as

pf = (-δ)

Cesos-Workshop-March-2006 8

Simplified quasistatic loadresponse model is applied

for initial rdquooptimization studyrdquo

kTot∙r = FE - FT

where is total linearized stiffness of mooring lines FE is external (low-frequency) excitation and FT is thruster force

Conversely r = (FE ndash FT)kTot

Cesos-Workshop-March-2006 9

Two types of loss functions are considered

bull Typical LQG type of loss function

L( r ) = KTFT2 + KFr2

(r is response FT is thruster force KT and KF are constants)

bull Loss function based on failure probability

L( r ) = KT FT2 + KPФ(-δ)

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 6: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 6

Definition of delta index

(ie due to waves)

is the mean breaking strength of the line

is the standard deviation of the breaking strength

Cesos-Workshop-March-2006 7

Computation of reliability index

bull Failure probability (pf) is probability that the extreme dynamic response will exceed critical level within a given reference duration

bull Failure probability is estimated for a stationary reference time interval of eg 20 minutes by application of a Gumbel distribution

bull Simplified relationship between delta-index and failure probability is expressed as

pf = (-δ)

Cesos-Workshop-March-2006 8

Simplified quasistatic loadresponse model is applied

for initial rdquooptimization studyrdquo

kTot∙r = FE - FT

where is total linearized stiffness of mooring lines FE is external (low-frequency) excitation and FT is thruster force

Conversely r = (FE ndash FT)kTot

Cesos-Workshop-March-2006 9

Two types of loss functions are considered

bull Typical LQG type of loss function

L( r ) = KTFT2 + KFr2

(r is response FT is thruster force KT and KF are constants)

bull Loss function based on failure probability

L( r ) = KT FT2 + KPФ(-δ)

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 7: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 7

Computation of reliability index

bull Failure probability (pf) is probability that the extreme dynamic response will exceed critical level within a given reference duration

bull Failure probability is estimated for a stationary reference time interval of eg 20 minutes by application of a Gumbel distribution

bull Simplified relationship between delta-index and failure probability is expressed as

pf = (-δ)

Cesos-Workshop-March-2006 8

Simplified quasistatic loadresponse model is applied

for initial rdquooptimization studyrdquo

kTot∙r = FE - FT

where is total linearized stiffness of mooring lines FE is external (low-frequency) excitation and FT is thruster force

Conversely r = (FE ndash FT)kTot

Cesos-Workshop-March-2006 9

Two types of loss functions are considered

bull Typical LQG type of loss function

L( r ) = KTFT2 + KFr2

(r is response FT is thruster force KT and KF are constants)

bull Loss function based on failure probability

L( r ) = KT FT2 + KPФ(-δ)

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 8: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 8

Simplified quasistatic loadresponse model is applied

for initial rdquooptimization studyrdquo

kTot∙r = FE - FT

where is total linearized stiffness of mooring lines FE is external (low-frequency) excitation and FT is thruster force

Conversely r = (FE ndash FT)kTot

Cesos-Workshop-March-2006 9

Two types of loss functions are considered

bull Typical LQG type of loss function

L( r ) = KTFT2 + KFr2

(r is response FT is thruster force KT and KF are constants)

bull Loss function based on failure probability

L( r ) = KT FT2 + KPФ(-δ)

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 9: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 9

Two types of loss functions are considered

bull Typical LQG type of loss function

L( r ) = KTFT2 + KFr2

(r is response FT is thruster force KT and KF are constants)

bull Loss function based on failure probability

L( r ) = KT FT2 + KPФ(-δ)

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 10: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 10

Two different types of control schemes are considered

bull PID control scheme

where e here is e= (rTarget ndash rstatic passive) = (rTarget - FEkTot)

which (by neglecting second and last term) simplifies into

FT = Kp (rTarget ndash rstatic passive) = Kp (rTarget - FEkTot)

bull LQG control scheme FT = -Cr

Normalized control factor is xc= CkTot

t

0d)(eiKedKepKTF

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 11: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 11

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 12: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 12

First type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 13: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 13

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 14: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 14

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 15: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 15

Second type of loss function versus vessel offset PID type of control scheme (KT∙kTot

2)KF = 01 (intermediate value) and FEkTot = 20

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 16: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 16

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 17: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 17

First type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 18: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 18

Second type of loss function expressed in terms of normalized control variable -LQG type of control scheme (KT∙kTot

2)KF = 10 and FEkTot = 20

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 19: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 19

Second type of loss function expressed in terms of normalized control variable - LQG type of control scheme (KT∙kTot

2)KF = 001 and FEkTot = 20

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 20: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 20

Comparison of optimal offsets for different loss functions

Coefficient ratio

(KT∙k Tot2)KF

and

(KT∙k Tot2)KP

Quadratic loss

function

Loss function

based on failure

probability

100 100m 200m

001 002m 000m

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 21: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 21

Example Position control of turret moored vessel

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 22: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 22

Vessel data

bull Length of vessel 175m

bull Beam 254m

bull Draught 95m

bull Displaced volume 24 140m3

bull Mooring lines are composed of a mixture of chains and wire lines

bull Representative linearized stiffness of the mooring system is 15∙104 Nm

bull Mean value of breaking strength of single line is 1128∙106 N

bull Standard deviation of the breaking strength is 75 of the mean value

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 23: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 23

Numerical simulation model

τ

is the mooring force

is the thruster force

wJp

bJτg(ηDννM

2

T

ψ

ψ)

)g(η

η = [pT ψ]T = [x y ψ]T is the position and heading in earth-fixed coordinates

ν = [wT ρ]T = [u v ρ]T is the translational and rotational velocities in body-fixed coordinates

M

is the inertia matrix D is the hydrodynamic damping matrix

b is a slowly varying bias term representing external forces due to wind currents and waves

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 24: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 24

Feedback control law based on back-stepping technique

bψJηgDνMςτ T

2

(20)

I

1

j sj

j jj s j s j s

b j j b j

s

r

T T

r

T2

w

w S

j

jT2 r

ppJ

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 25: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 25

λ γ and κ are strictly positive constants

rj is the length of the horizontal projection of mooring line number j

Tjrsquois the linearized mooring line tension in line j

pj is the horizontal position of the end-point at the anchor for the same mooring line

σbj is the standard deviation of the breaking strength of line number j

The target value of the reliability index is designated by δs

It can be shown that this controller is global exponentially stable

Notation

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 26: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 26

Time variation of water current velocity

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 27: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 27

Time variation of resultant environmental force

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 28: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 28

Time variation of vessel position in x-direction

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 29: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 29

Time variation of thruster force

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 30: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 30

Time variation of delta-index

In order for a delta-index of 44 to be optimal for the present

case study the ratio of (KT∙kTot2)KP needs to be 10-6 ie the

failure cost needs to be very high compared to the rdquounit thruster costrdquo

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm

Page 31: RELIABILITY-BASED STRUCTURAL OPTIMIZATION       FOR   POSITIONING OF MARINE VESSELS

Cesos-Workshop-March-2006 31

Summaryconclusions

bull A simplified model is applied in order to study optimal offset values (and corresponding values of the delta-index) when considering both the cost and reliability level

bull Two different loss functions are compared The first type is quadratic in the response while the second is proportional to the failure probability

bull It is demonstrated for a particular example how structural reliability criteria can be incorporated directly into the control algorithm