基于協(xié)同進化的多目標優(yōu)化算法研究

1、天津大學碩士學位論文基于協(xié)同進化的多目標優(yōu)化算法研究姓名：劉國興申請學位級別：碩士專業(yè)：系統(tǒng)工程指導教師：李敏強20080501ABSTRACTMulti—objectiVeOptimizationProbl鋤(MOP)hasbeenadi伍cultproblemaIldfocusforresearchiIlfieldofdecisionscieI】lce，Therealreadywerealotofclassicalmemodsfor

2、solVillgMOPsbeforeGeneticpdgorithms(GAs)wereappliedtosolveMOPsClassicalmultiobjectiVeoptimizationmethodshaVebeent110rougmydeVeloped，buttherearestiU10tsofshortcomillgsGAscouldh柚d1ela瑪eSpaceofproblemandgetacoupleofsolution

3、siIloneeVolutioIl，andGAsdonotneedmuchinfomationa_bouttheproblembeforestartiIlgtheoptimizationprocessHoweVer，thereisai11reconcilableconnictbe觚eenprema嘁conVe唱enceandlowconve瑪e11cespeedi11GAs，soGAshavelinlitsiIls01vinglli曲d

4、imellsion，multiIllodalMOPsCo—evolutionaUAlgorithm(CEA)，anew虹ndEVolutiona巧AlgorithmwMcharoseinl990’s，cansolVetheconflictmentionedaboVeperfectly，sousingCEAtos01VeMOPsisthe仃endintlle向tuleBasedoneXtensiVe粕dde印reViewofliteran

5、脫，athorou曲analysisaIldresearchonCEAsandMOPsispresented，themaillcontentsareasfollows：ThedeVelopmentofMOPsaIldtlleclassicalmethodsforsolvingMOPsissummarized，a11dtheliIllitsoftheclassicalme吐lodsarepointedoutThedeVelopmentof

6、GAsisrcViewed，andt11ebasicⅡ1eoriesa11dknowledgeofGAsaredetailedEValuationandcoInp撕sonofthe仃aditionalmetllodsforsolVingMOPswiththeGAsaremade，andtheiriIlspirationtootlleralgorit№sisrevealedThedeVelopmentof血eCEAsissu瑚marize

7、d，aIldthet11inkngofcooperatiVeandcompetitiVeCEAsisanalyzedTwoCEAs(cooperativeandcompetitiVe)tosolVeMOPsarepresented，aIldseVeraliIllproVememstrategiesareproposedtoaVoidsomeproblemsThesetwomemods鋤dMOGAareusedtosolve6teSt如n