資源描述：
《多目標優(yōu)化的粒子群算法及其應用研究》由會員上傳分享，免費在線閱讀，更多相關內容在學術論文-天天文庫。
1、大連理工大學碩士學位論文多目標優(yōu)化的粒子群算法及其應用研究姓名：陳紹新申請學位級別：碩士專業(yè)：軟件工程指導教師：周寬久20071118多目標優(yōu)化的粒子群算法及箕應用研究ResearchonMulti—ObjectiveOptimizationParticleSwarmAlgorithmanditsApplicationAbstractParticleSwarmOptimig缸icrnalgorithmisdevelopedinthepastfewyearstoresolvetheissueofmulti-objectiveop
2、timizationalgorithmintelligentgroup．Tbealgorithmisbasedonthisassumption：thevariousgroupsofparticlescallgeteffectivemessagefrompastexperienceandtheexperienceofotherparticles．Theexperimentswasfoundthatfor1hemajorityofoptimizationproblemsthisalgorithmhasmanyfeaturessuc
3、hasafasterconvergencerate，thelesstheneedforparametersandsoOILButitremainsinadequateinthedistributionandconvergenceofitssolutionset．Itsapplicationsal'eless．FromsomecurrentI它searckthispaperdesignedahybridimprovedParticleSwarmAlgorithmfromthreephrases，andappliedtotheNu
4、tridoncatemgComputingModel．11把workinthedissertationisshownasfollows：1．Summingupthemulti-objectiveoptimizationsolutionsofthetraditionalmethodsandevolutionaryalgofithmbased011thesolutionflgofithm,andfocusesolltheparticleswal'malgorithmanditsmulti-objeetiveopfimiz嘶onin
5、thefieIdofthestatusquo．2．Thispaperdesignedanewmulti-objectivePSOtoimprovethealgorithmfromthreephrases：(1)ItjoinedtabualgorithmandCrowdedmechanisminselectingoftheoptilnalvalueofthebest，sothatitavoidedgettingintotheoptimizinglocalityearlierandimprovedthesolutionprecis
6、ion．(2)Asolutionofalladvancedfitoessfunction,whichin打oducedtheconceptionoffeasiblezoneto自rcatwiththereswainedconditions．(3)Itputforwardthenon-dominatedset：astrategyabouttherandomchoiceinlabelswitching．Itimprovedthe富喃】曲瑚】e甩ciencyofthenon-dominatedsetandreducedthealgo
7、ritlma’stimecomplexity．3．ItdeignedanewnutritionCateringmodelwhichcanmeetallnutritionalneedsofthehumanbodyandtheneedsofthevarioustypeofthecrowd．他modelisspeciesdiversityandflexible．htheend,thenewnutritionCateringmodeIisdesignedandrealizedwiththencq～"multi-objectivePSO
8、accordingtothecharacteristics．Bycomparison、Ⅳiththetraditionalwaytosolvethemodelandexperimentalresults，itprovedfeasibilityandeffectiveness，
