基于遺傳算法的機會約束規(guī)劃區(qū)間估計.pdf

1、西南交通大學碩士學位論文基于遺傳算法的機會約束規(guī)劃區(qū)間估計姓名：陳高波申請學位級別：碩士專業(yè)：應用數學指導教師：劉海燕2003.3.1西南交通大學碩士研究生學位論文第1l頁AbstractStochasticprogrammingisanimportantbranchofmathematicalprogrammingHoweveritisverydifferentfromordinarymathematicalprogrammingHo

2、wtosolveitismuchmorecomplexthanthatofordinarymathematicalprogrammingbecauseofitsrandomcoefficientsThispapertriestostudystochastieprogramming，especiallychanceCOflstrfljnedprogrammingOffthetheoryofprobabilityandstatistiesA

3、tpresenttherearetwomainmethodstosolvingstochasticprogrammingThefirstistransformingitintoequivalentdeterministicprogrammingandthensolvingitbyusingthetheoryofdeterministicprogrammingthathasbeendevelopedsuccessfullyAnotheri

4、sgettingtheapproximateoptimumvalueandoptimumsolutionofchance—constrainedprogrammingthroughsomecertaingeneticalgorithmbasedonrandomsimulatedtechnologyThispapersummarizestwomethodsofchanceconstrainedprogrammingAccordingtot

5、hefirstmethod，thisarticlegeneralizesthecoefficienttypeofthosethatcanbetransformedintodeterministicprogrammingtoexponentialfamilyFormoregeneralchanceconstrainedprogramming。wehaveobtaineditsestimatedintervalandhavediscusse

6、dthefactorsthataffecttheprecisionofestimatedoptimumvalueandpointedoutthemethodthatraisesestimatedprecisionThisestimatedintervalcontainedanestimatedvalueofgoalfunctionInviewofthefactthatthegeneticalgorithmofstochasticprog

7、rammingbasedonrandomsimulatedtechnologyhassucceedgreatly，thispaperpointsoutthatchangingparametersofgeneticalgorithmcan曲tainasequenceofoptimumvaluesofgoalfunctionTakingthesegeneticalgorithmvaluesassamplingdata，wecangetfit

8、tingoptimumfunctionbyusingmultivariatesplineregressionandgettheLipschitzsconstantofthefittingoptimumfunctionSoforanychanceconstrainedprogrammingproblem，wecangetitsintervalestimatekeywords：StochasticProgramming，Chancecons