多目标优化的粒子群算法及其应用研究

《多目标优化的粒子群算法及其应用研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、大连理工大学硕士学位论文多目标优化的粒子群算法及其应用研究姓名：陈绍新申请学位级别：硕士专业：软件工程指导教师：周宽久20071118多目标优化的粒子群算法及箕应用研究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，