research article a forecasting model for feed grain...

Research ArticleA Forecasting Model for Feed Grain Demand Based onCombined Dynamic Model

Tiejun Yang Na Yang and Chunhua Zhu

School of Information Science and Engineering Henan University of Technology Zhengzhou 450001 China

Correspondence should be addressed to Chunhua Zhu zhuchunhuahauteducn

Received 6 April 2016 Revised 28 June 2016 Accepted 14 July 2016

Academic Editor Jorge Reyes

Copyright copy 2016 Tiejun Yang et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

In order to improve the long-term prediction accuracy of feed grain demand a dynamic forecast model of long-term feed graindemand is realized with joint multivariate regression model of which the correlation between the feed grain demand and itsinfluence factors is analyzed firstly then the change trend of various factors that affect the feed grain demand is predicted byusing ARIMAmodelThe simulation results show that the accuracy of proposed combined dynamic forecastingmodel is obviouslyhigher than that of the grey system model Thus it indicates that the proposed algorithm is effective

1 Introduction

The grain used in feeding is the second largest grain usedin China its quantity and proportion of the total grain con-sumption grow stably It is of great significance to ensure foodsecurity in our country by exploring the changes of feed graindemand and its influencing factors However the specialresearch of Chinarsquos feed grain demand is scattered whichlacks objective statistics and always exists in projections ofthe total grain consumptionThe forecasting methods of feedgrain demand in existing literature can be divided into twokinds one is using some quantitative methods such as timeseries regression model of consumer demand system andfarming grain consumption based on the analysis about thesituation of the feeding food consumption over the past fewyears to analysis and forecast [1 2] the other is from theperspective of nutrition standards analysis of meat eggsmilk per capita consumption of aquatic products to predictthe future demand for animal products and then use theratio of feed to meat (ie the conversion rate of feed grains)to predict the feed grain demand [3 4] Actually the feedgrain demand is affected by population growth urbanizationlevel per capita income (urban residents per capita incomeand rural ones per capita income) and other factors [56] which suggest that there should be a comprehensivesurvey about correlation degree between the feed grain

demand and its influence factors for improving the predictionaccuracy and the corresponding prediction model shouldbe generalized In this paper the correlation coefficients offeed grain demand and its influence factors are calculatedquantitatively on the basis of the second kind of forecastingmethod then the major factors have been chosen finallythe dynamic prediction of influence factors and feed graindemand can be realized by using the ARIMA model andmultiple regression model respectively

2 Relational Coefficient Analysis of InfluenceFactors to Feed Grain Demand

21 Grey Relational Analysis The essence of grey relationaldegree is to make a geometric comparison in the data serieswhich are responded to the changing characteristics of allfactors The closer the curves are the greater the relationalgrade of the corresponding series is and vice versaThe use ofthe grey relational analysis can define the changing trend ofall factors in this system and find out the main factors whichaffect the further development of the system so as to graspthe main features of things and the principal contradictionpromote and guide the system to rapid health and efficientdevelopment [7]The basic steps of grey relational analysis areas follows

Hindawi Publishing CorporationComputational Intelligence and NeuroscienceVolume 2016 Article ID 5329870 6 pageshttpdxdoiorg10115520165329870

2 Computational Intelligence and Neuroscience

Step 1 Assume that the reference sequence is 1199090(119896) and

related comparison sequences are 119909119894(119896) They are expressed

as 1199090(119896) = 119909

0(1) 1199090(2) 119909

0(119899) and

119909119894(119896) = 119909

119894(1) 119909

119894(2) 119909

119894(119899) (119894 = 1 2 119898) (1)

Step 2 Dis-dimension treatment to the data sequence [8]Here we illustrate the initiating Then it can get the refer-ence sequence 119910

0(119896) and comparison sequences 119910

119894(119896) (119894 =

1 2 119898 119896 = 1 2 119899)

Step 3 The absolute difference sequences Δ0119894(119896) between

reference sequence1199100(119896) and comparison sequences119910

119894(119896) are

calculated by the formula

Δ0119894(119896) =

10038161003816100381610038161199100(119896) minus 119910

119894(119896)1003816100381610038161003816

= Δ119894 (1) Δ 119894 (

2) Δ 119894 (119899)

(119894 = 1 2 119898)

(2)

Step 4 Identify the absolute maximum Δmax and minimumΔmin from absolute difference sequence

Step 5 Calculate the grey relational coefficient The formulais

1198710119894 (119896) =

(Δmax + Δmin)

(Δ0119894(119896) + Δmax)

(3)

Step 6 Calculate correlation degree

1198770119894(119896) =

1

119899

119899

sum

119896=1

1198710119894(119896)

=

1

119899

1198710119894(1) + 119871

0119894(2) + sdot sdot sdot + 119871

0119894(119899)

(4)

22 Prediction for the Feed Grain Demand by Using Multi-ple Linear Regression According to grey relational analysisthe domestic population urbanization level and per capitaincome of urban and rural residents are the main factorsaffecting the feed grain demand Based on the modelingprinciple of multiple regression model the linear regressionmodel of the feed grain demand is set up the structure formof the model [9]

1199100= 1198800+ 11988011199091+ 11988021199092+ 11988031199093+ 120576 120576 sim 119873 (0 120575

2) (5)

In the formula 1198801 1198802 and 119880

3are the undetermined

parameters (regression parameters) with 120576 for unobservablerandom error

23 Prediction for Main Factors That Influence the Feed GrainDemand The ARIMA model from literature is adopted topredict the change trend of impact factors [10] Suppose that120596119905is the predictive value in 119905 time of various influence factors

and 120596119905minus1 120596119905minus2 120596

119905minus119901are actual values of various impact

factors in past 119901 years Setting 120596119905= (1 minus 119871)

119889119910119905 among it 119910

119905

is a single integer sequence with 119889 order 120596119905is the stationary

Data input

Correlationcalculation

Multiple regressionforecast of demanding forfeeding grains

Impact factorsprediction

Figure 1 Dynamic prediction simulation process of feed graindemand

series [11] thus the general model of the ARMA model canbe expressed as

120596119905= 1205931120596119905minus1+ 1205932120596119905minus2+ sdot sdot sdot + 120593

119901120596119905minus119901+ 120576119905+ 1205791120576119905minus1+ sdot sdot sdot

+ 120579119902120576119905minus119902

(6)

In the formula 119901 and 119902 are respectively called autore-gressive order number and average order number Suppose 119871as the lag operator then

119871120596119905= 120596119905minus1

119871119901120596119905= 120596119905minus119901

(7)

Equation (6) can be rewritten as

120593 (119871) 120596119905= Θ (119871) 120576

119905 (8)

Among it 120593(119871) = 1 minus 1205931119871 minus 120593

21198712minus sdot sdot sdot minus 120593

119901119871119901 and Θ(119871) =

1 + 1205791119871 + 12057921198712+ sdot sdot sdot + 120579

119902119871119902

ARMA(119901 119902) model in formula (7) can be expressed asARIMA(119901 119889 119902) after 119889 order difference transformation

120593 (119871) (1 minus 119871)119889119910119905= Θ (119871) 120576

119905 (9)

120576119905is a white noise process with its mean value which is 0 and

variance is 1205902 [12]

3 Simulation Analysis

The dynamic simulation process based on the ARIMAmodeland multiple regression model to predict feed grain demandis shown in Figure 1

The dynamic prediction algorithm of feed grain demandis shown in Figure 1 define the year of 1981 as 119905 = 1 andthus 2007 as 119905 = 27 The feed grain demand of urban andrural population is respectively expressed as 119910

0(119905) and 119910

1(119905)

the three factors are respectively defined as 1199091(119905) 1199092(119905) and

1199093(119905) According to the simulation process shown in Figure 1

the forecast process of feed grain demand in this paper isshown in the following

Computational Intelligence and Neuroscience 3

(1) When t = 1sim27 calculate the correlation degree andrelational sequence respectively between 119910

0(119905) and

1199101(119905) and 119909

1(119905) 1199092(119905) and 119909

3(119905)

(2) Use ARIMA model to predict 1199091(119905) 1199092(119905) and 119909

3(119905)

when 119905 gt 27(3) Use multiple regressionmethod to predict urban feed

grain demand 1199100(119905) (119905 = 28) and rural feed grain

demand 1199101(119905) (119905 = 28)

(4) Repeat (3) Urban and rural long-term prediction offeed grain demand can be completed

31 Correlation Calculation The data about the feed graindemand urban and rural population urbanization level andurban and rural residents per capita income between 1981 and2007 are selected fromRural China Statistical Yearbook [13] asthe training data meanwhile the data from 2008 to 2012 areselected as the precision test data as shown in Table 1The feedgrain demand can be got by the sum of per capita meat eggmilk and aquatic product consumption multiplied by theurban and rural population respectively and then accordingto the conversion ratio of feed grain to meat which is 37 to 1the conversion ratio to egg which is 27 to 1 the conversionratio to milk which is 05 to 1 and the conversion ratio toaquatic material which is 04 to 1 to get the final result [14 15]

The correlation degree and relational order are obtainedby using the grey correlation analysis method while thedata about the feed grain demand are calculated in Table 1as reference sequence at the same time urban and ruralpopulation urbanization level and urban and rural residentsper capita income are calculated as comparative sequenceThe results are shown in Table 2

As shown in Table 2 the correlation degree and relationalorder of various factors which affected the urban and ruralfeed grain demand are not completely the same on the basisof that it will be able to improve the prediction accuracy bypredicting towns and rural feed grain demand separately

32 Impact Factors Prediction ARIMA(119901 119889 119902) modeldescribed in Section 23 is adopted to predict the threefactors including urban and rural population urbanizationlevel and urban and rural residents per capita income Theprediction of impact factors for urban feed grain demand in2008 is taken as an example in this paper and the results areshown in Table 3 The forecast data will be used to forecastfeed grain demand in 2008

33 Prediction for Feed Grain Demand by Using MultipleRegression Themultiple regressionmodel of urban and ruraldemand for feed grain demands is set up respectively in2008 by usingEVIEWS statistical software while three factorsmentioned above are taken as independent variables andChinarsquos urban and rural residentsrsquo feed grain demand is takenas the dependent variable The models are shown as follows

1199100= minus4240163 + 15153119909

01+ 2104194119909

02

minus 195000611990903

(10)

1199101= minus21283643 + 232867119909

11+ 3920705119909

12

minus 54332811990913

(11)

Among them 1199100and 119910

1represent the urban and rural

feed grain demand respectively 11990901is urban population and

11990911

is rural population 11990902

and 11990912

represent urbanizationlevel 119909

03is urban residents per capita income and 119909

13is

rural residents per capita incomeThepredicted value of threefactors in 2008 was typed in (10) and (11) respectively thenthe value of urban and rural feed grain demand in 2008 can becalculated the results are 9807134 tons and 66637249 tons

In the above multivariate regression model of urban andrural feed grain demand the model prediction coefficientof different years will change dynamically as the change ofcorrelation of feed grains and affecting factors then it formsa dynamic forecast system

34 Simulation Results The value of feed grain demand in2008ndash2012 can be predicted according to (10) and (11) theresult is shown in Table 4 A grey forecasting model byusing residual error correction on the feed grain demand inliterature [16] is also given in Table 4

From Table 4 and combined with the feed grain demandbetween urban and rural areas since 1981 it can be seen thatthe basic trend of feed grain demand overall present risessteadily [17 18] The feed grain demand increased by 4 timesand the average annual growth rate is 148 from 1981 to2007 Analysis shows that the income level of our countryresidents is low and the consumption structure is unitarymainly grain consumption before the reform and open policyIn recent years the demand for animal products structure ischanging and it mainly displays in the increasing demand formeat eggs milk and aquatic products because peoplersquos livingstandards have been continuously improved

In addition compared with the grey system model inliterature [16] the joint dynamic prediction model in thispaper can track the change of impact factors so it can achievegood long-term forecasts Meanwhile the mean relative errorof proposed model is 046 and has higher superiority inforecasting precision compared with traditional grey fore-casting model of which the mean relative error is 64 Itis fully illustrated that the dynamic impact factor regressionanalysis method used to predict the feed grain demand isfeasible

4 Conclusion

The dynamic influence factors in combination with multi-variate regression analysis method are used in this paperto forecast the feed grain demand in China since 2008Prediction results show that Chinarsquos demand for feed grainswill increase year by year in the next 10 years and the averagerelative error between the actual and predicted value byusing the dynamic impact factor regression model is 046superior to the traditional grey system model At presentChinarsquos feed grain demand represents more than 30 of thetotal demand for grain the proportion of which feed graindemand on total demand for grain increased year by yearshows the increasing influence of feed grains on food security


Table1Statisticaldataof

vario

usim

pactfactors

Year

Meat

Egg

Milk

Aquatic

prod

uct

Popu

latio

n(te

nthou

sand

peop

le)

Perc

apita

income

(yuan)

Urbanizationlevel

()

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

1981

205

9452

1341

07

7313

20171

79901

5004

2234

142

1982

2199

59

1445

07

7713

21480

80174

5353

2701

144

1983

225

108

69

1646

08

81

1622274

80734

5646

3098

146

1984

228

115

7618

52

08

7817

24017

80340

6521

3553

147

1985

2412

88

21

64

08

7816

25094

80757

7391

3976

148

1986

253

129

7121

47

1482

1926366

81141

9009

4238

151987

254

129

66

23

55

1179

227674

81626

10021

4626

151

1988

237

1269

23

51

1171

1928661

82365

11802

5449

153

1989

239

123

7124

42

176

21

29540

83164

13739

6015

154

1990

252

126

7324

46

1177

21

30195

84138

15102

6863

155

1991

266

135

83

27

47

138

22

31203

84620

17006

7086

159

1992

265

133

9529

55

1582

23

32175

84996

20266

784

162

1993

26133

89

29

54

09

828

33173

85344

25774

9216

165

1994

243

126

973

53

07

85

334169

85681

34962

1221

168

1995

236

131

9732

46

06

9234

35174

85947

4283

15777

172

1996

258

148

9634

48

08

925

37

37304

85085

48389

19261

184

1997

255

151

111

41

51

193

38

3944

984177

51603

20901

196

1998

255

155

102

41

62

09

984

37

41608

83153

54251

2162

208

1999

267

164

109

43

791

103

38

43748

82038

5854

22103

222000

254

183

112

48

9911

117

39

45906

80837

6280

22534

232

2001

265

182

104

47

119

121033

41

4806

479563

68596

23664

244

2002

325

186

106

47

157

12132

44

50212

78241

77028

24756

258

2003

329

197

112

48

186

17134

47

52376

76851

84722

26222

272

2004

293

192

104

46

188

2125

45

54283

75705

94216

29364

289

2005

329

224

104

47

179

29

126

49

56212

74544

10493

32549

307

2006

321

223

104

5183

31

135

58288

73160

117595

3587

325

2007

318

205

103

47

178

35

142

54

60633

71496

137858

4140

4343

2008

312

202

107

54

152

34

119

52

62403

70399

157808

47606

362009

347

215

106

53

149

36

122

53

64512

68938

171747

51532

377

2010

347

222

1051

1436

152

52

66978

67113

191094

5919

388

2011

352

233

101

54

137

52

146

54

69079

65656

218098

69773

406

2012

357

235

105

59

1453

152

54

71182

64222

245647

79166

424

Note(1)u

nitperc

apita

consum

ptionin

kilogram

s(2)the

dataarefrom

RuralC

hina

Statistica

lYearbook


Table 2 The grey correlation analysis about each influencing factor in 1981ndash2007 of urban and rural feed grain demand

Influencing factor Urban RuralCorrelation degree Relational order Correlation degree Relational order

(Urbanrural) population 09370 1 09255 2Urbanization level 09047 2 09641 1(Urbanrural) per capita income 07236 3 06881 3

Table 3 Predicted value of various influencing factors in 2008

Influencing factor Model Adjusted 1198772 Predicted value in 2008Urban population ARIMA(3 2 6) 0956 6296545Urbanization level ARIMA(7 2 2) 0848 361254Urban residents per capita income ARIMA(4 2 5) 0876 1587219

Table 4 The comparison between the actual value and predicted value of feed grain demand under different prediction models (unit tenthousand tons)

Year Actual value Predicted value Relative error Mean relative error

Combined dynamic forecasting model

2008 163321 164709 08

0462009 176652 177953 072010 179810 179284 022011 186872 187985 052012 192676 192438 01

Grey forecasting model

2008 163321 179254 97

642009 176652 184269 432010 179810 192467 702011 186872 198751 642012 192676 201448 46

so it has become a necessary work to research the feed graindemand deeply for ensuring food security

Competing Interests

The authors declare that they have no competing interests

Acknowledgments

This research was financially supported by the National FoodIndustry Commonweal Special Scientific Research Projects(no 201413001)

References

[1] T Weiming and J Chudleigh Chinarsquos Feed Grain Marketdevelopment and Prospect AARCWorking Paper Series 1998

[2] X Yu and D Abler ldquoThe demand for food quality in ruralChinardquo American Journal of Agricultural Economics vol 91 no1 pp 57ndash69 2009

[3] L R Brown Who Will Feed China Wake up Call for a SmallPlanet WorldWatch Norton and Co New York NY USA 1995

[4] M Gao Q Luo Y Liu and J Mi ldquoGrain consumptionforecasting in China for 2030 and 2050 volume and vari-etiesrdquo in Proceedings of the 3rd International Conference onAgro-Geoinformatics (Agro-Geoinformatics rsquo14) pp 1ndash6 BeijingChina August 2014

[5] X Xin W Tian and Z Zhou ldquoChanging patterns of feed grainproduction andmarketing in Chinardquo Agribusiness PerspectivesPaper 47 2001

[6] NMinot and F Goletti ldquoRicemarket liberalization and povertyin Vietnamrdquo IFPRI Research Report 114 IFPRI WashingtonDC USA 2000

[7] M Hao and L Xiang ldquoGrey relational analysis for impact fac-tors of micro-milling surface roughnessrdquo in Proceedings of the12th IEEE International Conference on Electronic Measurementamp Instruments (ICEMI rsquo15) pp 109ndash113 IEEE Qingdao ChinaJuly 2015

[8] R Sallehuddin S M H Shamsuddin and S Z Mohd HashimldquoApplication of grey relational analysis for multivariate timeseriesrdquo in Proceedings of the 8th International Conference onIntelligent Systems Design and Applications (ISDA rsquo08) pp 432ndash437 Kaohsiung Taiwan November 2008

[9] Q Wang F Xia and X Wang ldquoIntegration of grey model andmultiple regression model to predict energy consumptionrdquo inProceedings of the International Conference on Energy and Envi-ronment Technology (ICEET rsquo09) pp 194ndash197 Guilin ChinaOctober 2009

[10] T Yang N Yang andC Zhu ldquoInvestigation of grain output pre-diction based on ARIMA modelrdquo Journal of Henan Universityof Technology (Natural Science Edition) vol 36 no 5 pp 24ndash272015

[11] K K Suresh and S R Krishna Priya ldquoForecasting sugarcaneyield of Tamilnadu using ARIMA modelsrdquo Sugar Tech vol 13no 1 pp 23ndash26 2011


[12] X Xu L Cao J Zhou and F Su ldquoStudy and application of grainyield forecasting modelrdquo in Proceedings of the 4th InternationalConference on Computer Science and Network Technology (ICC-SNT rsquo15) pp 652ndash656 Harbin China December 2015

[13] National Bureau of Statistics of China China Rural StatisticalYearbook China Statistics Press Beijing China 2014

[14] J van Zyl Grain Worth Gold ldquoFeeding Animals IncreasesDemand and Pushes up Pricesrdquo Business Strategy 2007

[15] F Qin and X Chen Chinese Farmersrsquo Food ConsumptionResearch China Agriculture Press Beijing China 2007

[16] M Li and H Lei Study on Chinarsquos food security in the new situ-ation [PhD thesis] Huazhong Agricultural University WuhanChina 2005

[17] C Aubert ldquoFood security and consumption patterns in Chinathe grain problemrdquoChina Perspectives vol 2008 no 2 pp 5ndash232014

[18] X Yu and D Abler ldquoThe demand for food quality in RuralChinardquo American Journal of Agricultural Economics vol 91 no1 pp 57ndash69 2009

Submit your manuscripts athttpwwwhindawicom

Computer Games Technology

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Distributed Sensor Networks


Advances in

FuzzySystems

Hindawi Publishing Corporationhttpwwwhindawicom

Volume 2014


ReconfigurableComputing

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014


Applied Computational Intelligence and Soft Computing

thinspAdvancesthinspinthinsp

Artificial Intelligence

HindawithinspPublishingthinspCorporationhttpwwwhindawicom Volumethinsp2014

Advances inSoftware EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Electrical and Computer Engineering

Journal of

Journal of

Computer Networks and Communications


Hindawi Publishing Corporation

httpwwwhindawicom Volume 2014

Advances in

Multimedia


Biomedical Imaging


ArtificialNeural Systems

Advances in


RoboticsJournal of



Computational Intelligence and Neuroscience

Industrial EngineeringJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Human-ComputerInteraction

Advances in

Computer EngineeringAdvances in



Step 1 Assume that the reference sequence is 1199090(119896) and

related comparison sequences are 119909119894(119896) They are expressed

as 1199090(119896) = 119909

0(1) 1199090(2) 119909

0(119899) and

119909119894(119896) = 119909

119894(1) 119909

119894(2) 119909

119894(119899) (119894 = 1 2 119898) (1)

Step 2 Dis-dimension treatment to the data sequence [8]Here we illustrate the initiating Then it can get the refer-ence sequence 119910

0(119896) and comparison sequences 119910

119894(119896) (119894 =

1 2 119898 119896 = 1 2 119899)

Step 3 The absolute difference sequences Δ0119894(119896) between

reference sequence1199100(119896) and comparison sequences119910

119894(119896) are

calculated by the formula

Δ0119894(119896) =

10038161003816100381610038161199100(119896) minus 119910

119894(119896)1003816100381610038161003816

= Δ119894 (1) Δ 119894 (

2) Δ 119894 (119899)

(119894 = 1 2 119898)

(2)

Step 4 Identify the absolute maximum Δmax and minimumΔmin from absolute difference sequence

Step 5 Calculate the grey relational coefficient The formulais

1198710119894 (119896) =

(Δmax + Δmin)

(Δ0119894(119896) + Δmax)

(3)

Step 6 Calculate correlation degree

1198770119894(119896) =

1

119899

119899

sum

119896=1

1198710119894(119896)

=

1

119899

1198710119894(1) + 119871

0119894(2) + sdot sdot sdot + 119871

0119894(119899)

(4)

22 Prediction for the Feed Grain Demand by Using Multi-ple Linear Regression According to grey relational analysisthe domestic population urbanization level and per capitaincome of urban and rural residents are the main factorsaffecting the feed grain demand Based on the modelingprinciple of multiple regression model the linear regressionmodel of the feed grain demand is set up the structure formof the model [9]

1199100= 1198800+ 11988011199091+ 11988021199092+ 11988031199093+ 120576 120576 sim 119873 (0 120575

2) (5)

In the formula 1198801 1198802 and 119880

3are the undetermined

parameters (regression parameters) with 120576 for unobservablerandom error

23 Prediction for Main Factors That Influence the Feed GrainDemand The ARIMA model from literature is adopted topredict the change trend of impact factors [10] Suppose that120596119905is the predictive value in 119905 time of various influence factors

and 120596119905minus1 120596119905minus2 120596

119905minus119901are actual values of various impact

factors in past 119901 years Setting 120596119905= (1 minus 119871)

119889119910119905 among it 119910

119905

is a single integer sequence with 119889 order 120596119905is the stationary

Data input

Correlationcalculation

Multiple regressionforecast of demanding forfeeding grains

Impact factorsprediction

Figure 1 Dynamic prediction simulation process of feed graindemand

series [11] thus the general model of the ARMA model canbe expressed as

120596119905= 1205931120596119905minus1+ 1205932120596119905minus2+ sdot sdot sdot + 120593

119901120596119905minus119901+ 120576119905+ 1205791120576119905minus1+ sdot sdot sdot

+ 120579119902120576119905minus119902

(6)

In the formula 119901 and 119902 are respectively called autore-gressive order number and average order number Suppose 119871as the lag operator then

119871120596119905= 120596119905minus1

119871119901120596119905= 120596119905minus119901

(7)

Equation (6) can be rewritten as

120593 (119871) 120596119905= Θ (119871) 120576

119905 (8)

Among it 120593(119871) = 1 minus 1205931119871 minus 120593

21198712minus sdot sdot sdot minus 120593

119901119871119901 and Θ(119871) =

1 + 1205791119871 + 12057921198712+ sdot sdot sdot + 120579

119902119871119902

ARMA(119901 119902) model in formula (7) can be expressed asARIMA(119901 119889 119902) after 119889 order difference transformation

120593 (119871) (1 minus 119871)119889119910119905= Θ (119871) 120576

119905 (9)

120576119905is a white noise process with its mean value which is 0 and

variance is 1205902 [12]

3 Simulation Analysis

The dynamic simulation process based on the ARIMAmodeland multiple regression model to predict feed grain demandis shown in Figure 1

The dynamic prediction algorithm of feed grain demandis shown in Figure 1 define the year of 1981 as 119905 = 1 andthus 2007 as 119905 = 27 The feed grain demand of urban andrural population is respectively expressed as 119910

0(119905) and 119910

1(119905)

the three factors are respectively defined as 1199091(119905) 1199092(119905) and

1199093(119905) According to the simulation process shown in Figure 1

the forecast process of feed grain demand in this paper isshown in the following



0(119905) and

1199101(119905) and 119909

1(119905) 1199092(119905) and 119909

3(119905)


3(119905)



demand 1199101(119905) (119905 = 28)







1199100= minus4240163 + 15153119909

01+ 2104194119909

02

minus 195000611990903

(10)

1199101= minus21283643 + 232867119909

11+ 3920705119909

12

minus 54332811990913

(11)




11990911


and 11990912



13is






4 Conclusion




vario

usim

pactfactors

Year

Meat

Egg

Milk

Aquatic

prod

uct

Popu

latio

n(te

nthou

sand

peop

le)

Perc

apita

income

(yuan)

Urbanizationlevel

()

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

1981

205

9452

1341

07

7313

20171

79901

5004

2234

142

1982

2199

59

1445

07

7713

21480

80174

5353

2701

144

1983

225

108

69

1646

08

81

1622274

80734

5646

3098

146

1984

228

115

7618

52

08

7817

24017

80340

6521

3553

147

1985

2412

88

21

64

08

7816

25094

80757

7391

3976

148

1986

253

129

7121

47

1482

1926366

81141

9009

4238

151987

254

129

66

23

55

1179

227674

81626

10021

4626

151

1988

237

1269

23

51

1171

1928661

82365

11802

5449

153

1989

239

123

7124

42

176

21

29540

83164

13739

6015

154

1990

252

126

7324

46

1177

21

30195

84138

15102

6863

155

1991

266

135

83

27

47

138

22

31203

84620

17006

7086

159

1992

265

133

9529

55

1582

23

32175

84996

20266

784

162

1993

26133

89

29

54

09

828

33173

85344

25774

9216

165

1994

243

126

973

53

07

85

334169

85681

34962

1221

168

1995

236

131

9732

46

06

9234

35174

85947

4283

15777

172

1996

258

148

9634

48

08

925

37

37304

85085

48389

19261

184

1997

255

151

111

41

51

193

38

3944

984177

51603

20901

196

1998

255

155

102

41

62

09

984

37

41608

83153

54251

2162

208

1999

267

164

109

43

791

103

38

43748

82038

5854

22103

222000

254

183

112

48

9911

117

39

45906

80837

6280

22534

232

2001

265

182

104

47

119

121033

41

4806

479563

68596

23664

244

2002

325

186

106

47

157

12132

44

50212

78241

77028

24756

258

2003

329

197

112

48

186

17134

47

52376

76851

84722

26222

272

2004

293

192

104

46

188

2125

45

54283

75705

94216

29364

289

2005

329

224

104

47

179

29

126

49

56212

74544

10493

32549

307

2006

321

223

104

5183

31

135

58288

73160

117595

3587

325

2007

318

205

103

47

178

35

142

54

60633

71496

137858

4140

4343

2008

312

202

107

54

152

34

119

52

62403

70399

157808

47606

362009

347

215

106

53

149

36

122

53

64512

68938

171747

51532

377

2010

347

222

1051

1436

152

52

66978

67113

191094

5919

388

2011

352

233

101

54

137

52

146

54

69079

65656

218098

69773

406

2012

357

235

105

59

1453

152

54

71182

64222

245647

79166

424

Note(1)u

nitperc

apita

consum

ptionin

kilogram

s(2)the

dataarefrom

RuralC

hina

Statistica

lYearbook










2008 163321 164709 08

0462009 176652 177953 072010 179810 179284 022011 186872 187985 052012 192676 192438 01


2008 163321 179254 97

642009 176652 184269 432010 179810 192467 702011 186872 198751 642012 192676 201448 46


Competing Interests


Acknowledgments


References



























Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





0(119905) and

1199101(119905) and 119909

1(119905) 1199092(119905) and 119909

3(119905)


3(119905)



demand 1199101(119905) (119905 = 28)







1199100= minus4240163 + 15153119909

01+ 2104194119909

02

minus 195000611990903

(10)

1199101= minus21283643 + 232867119909

11+ 3920705119909

12

minus 54332811990913

(11)




11990911


and 11990912



13is






4 Conclusion




vario

usim

pactfactors

Year

Meat

Egg

Milk

Aquatic

prod

uct

Popu

latio

n(te

nthou

sand

peop

le)

Perc

apita

income

(yuan)

Urbanizationlevel

()

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

1981

205

9452

1341

07

7313

20171

79901

5004

2234

142

1982

2199

59

1445

07

7713

21480

80174

5353

2701

144

1983

225

108

69

1646

08

81

1622274

80734

5646

3098

146

1984

228

115

7618

52

08

7817

24017

80340

6521

3553

147

1985

2412

88

21

64

08

7816

25094

80757

7391

3976

148

1986

253

129

7121

47

1482

1926366

81141

9009

4238

151987

254

129

66

23

55

1179

227674

81626

10021

4626

151

1988

237

1269

23

51

1171

1928661

82365

11802

5449

153

1989

239

123

7124

42

176

21

29540

83164

13739

6015

154

1990

252

126

7324

46

1177

21

30195

84138

15102

6863

155

1991

266

135

83

27

47

138

22

31203

84620

17006

7086

159

1992

265

133

9529

55

1582

23

32175

84996

20266

784

162

1993

26133

89

29

54

09

828

33173

85344

25774

9216

165

1994

243

126

973

53

07

85

334169

85681

34962

1221

168

1995

236

131

9732

46

06

9234

35174

85947

4283

15777

172

1996

258

148

9634

48

08

925

37

37304

85085

48389

19261

184

1997

255

151

111

41

51

193

38

3944

984177

51603

20901

196

1998

255

155

102

41

62

09

984

37

41608

83153

54251

2162

208

1999

267

164

109

43

791

103

38

43748

82038

5854

22103

222000

254

183

112

48

9911

117

39

45906

80837

6280

22534

232

2001

265

182

104

47

119

121033

41

4806

479563

68596

23664

244

2002

325

186

106

47

157

12132

44

50212

78241

77028

24756

258

2003

329

197

112

48

186

17134

47

52376

76851

84722

26222

272

2004

293

192

104

46

188

2125

45

54283

75705

94216

29364

289

2005

329

224

104

47

179

29

126

49

56212

74544

10493

32549

307

2006

321

223

104

5183

31

135

58288

73160

117595

3587

325

2007

318

205

103

47

178

35

142

54

60633

71496

137858

4140

4343

2008

312

202

107

54

152

34

119

52

62403

70399

157808

47606

362009

347

215

106

53

149

36

122

53

64512

68938

171747

51532

377

2010

347

222

1051

1436

152

52

66978

67113

191094

5919

388

2011

352

233

101

54

137

52

146

54

69079

65656

218098

69773

406

2012

357

235

105

59

1453

152

54

71182

64222

245647

79166

424

Note(1)u

nitperc

apita

consum

ptionin

kilogram

s(2)the

dataarefrom

RuralC

hina

Statistica

lYearbook










2008 163321 164709 08

0462009 176652 177953 072010 179810 179284 022011 186872 187985 052012 192676 192438 01


2008 163321 179254 97

642009 176652 184269 432010 179810 192467 702011 186872 198751 642012 192676 201448 46


Competing Interests


Acknowledgments


References



























Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





vario

usim

pactfactors

Year

Meat

Egg

Milk

Aquatic

prod

uct

Popu

latio

n(te

nthou

sand

peop

le)

Perc

apita

income

(yuan)

Urbanizationlevel

()

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

Urban

Rural

1981

205

9452

1341

07

7313

20171

79901

5004

2234

142

1982

2199

59

1445

07

7713

21480

80174

5353

2701

144

1983

225

108

69

1646

08

81

1622274

80734

5646

3098

146

1984

228

115

7618

52

08

7817

24017

80340

6521

3553

147

1985

2412

88

21

64

08

7816

25094

80757

7391

3976

148

1986

253

129

7121

47

1482

1926366

81141

9009

4238

151987

254

129

66

23

55

1179

227674

81626

10021

4626

151

1988

237

1269

23

51

1171

1928661

82365

11802

5449

153

1989

239

123

7124

42

176

21

29540

83164

13739

6015

154

1990

252

126

7324

46

1177

21

30195

84138

15102

6863

155

1991

266

135

83

27

47

138

22

31203

84620

17006

7086

159

1992

265

133

9529

55

1582

23

32175

84996

20266

784

162

1993

26133

89

29

54

09

828

33173

85344

25774

9216

165

1994

243

126

973

53

07

85

334169

85681

34962

1221

168

1995

236

131

9732

46

06

9234

35174

85947

4283

15777

172

1996

258

148

9634

48

08

925

37

37304

85085

48389

19261

184

1997

255

151

111

41

51

193

38

3944

984177

51603

20901

196

1998

255

155

102

41

62

09

984

37

41608

83153

54251

2162

208

1999

267

164

109

43

791

103

38

43748

82038

5854

22103

222000

254

183

112

48

9911

117

39

45906

80837

6280

22534

232

2001

265

182

104

47

119

121033

41

4806

479563

68596

23664

244

2002

325

186

106

47

157

12132

44

50212

78241

77028

24756

258

2003

329

197

112

48

186

17134

47

52376

76851

84722

26222

272

2004

293

192

104

46

188

2125

45

54283

75705

94216

29364

289

2005

329

224

104

47

179

29

126

49

56212

74544

10493

32549

307

2006

321

223

104

5183

31

135

58288

73160

117595

3587

325

2007

318

205

103

47

178

35

142

54

60633

71496

137858

4140

4343

2008

312

202

107

54

152

34

119

52

62403

70399

157808

47606

362009

347

215

106

53

149

36

122

53

64512

68938

171747

51532

377

2010

347

222

1051

1436

152

52

66978

67113

191094

5919

388

2011

352

233

101

54

137

52

146

54

69079

65656

218098

69773

406

2012

357

235

105

59

1453

152

54

71182

64222

245647

79166

424

Note(1)u

nitperc

apita

consum

ptionin

kilogram

s(2)the

dataarefrom

RuralC

hina

Statistica

lYearbook










2008 163321 164709 08

0462009 176652 177953 072010 179810 179284 022011 186872 187985 052012 192676 192438 01


2008 163321 179254 97

642009 176652 184269 432010 179810 192467 702011 186872 198751 642012 192676 201448 46


Competing Interests


Acknowledgments


References



























Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in












2008 163321 164709 08

0462009 176652 177953 072010 179810 179284 022011 186872 187985 052012 192676 192438 01


2008 163321 179254 97

642009 176652 184269 432010 179810 192467 702011 186872 198751 642012 192676 201448 46


Competing Interests


Acknowledgments


References



























Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in


















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in



research article a forecasting model for feed grain...

Documents