欢迎来到天天文库
浏览记录
ID:32051361
大小:9.90 MB
页数:61页
时间:2019-01-31
《高校动态多目标优化算法与其在pid控制中的应用分析》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、硕士学位论文AbstractInengineeringdesign,real-timecontrol,artificialintelligenceandotherengineeringfields,thereareawiderangeofoptimizationproblemsthattheobjectivesandconstraintswillchangeovertime.Andthiskindofproblemiscalleddynamicoptimizationproblem.Iftheoptimizationpro
2、blemhasmultipleconflictingoptimizationgoals,thenitisknownasadynamicmulti-objectiveoptimizationproblem.Thecomplexnatureofthedynamicmulti-objectiveoptimizationproblemitselfmakesthiskindofproblemverydifficulttosolve.Itrequiresthealgorithmtomakeadjustmentsaccordingtoc
3、hangesintheenvironmentcontinuously,andsearchthechangedParetooptimalsolutionsetasquicklyaspossible.Thiswillinevitablyrequiredynamicmulti—objectiveoptimizationalgorithmhashighsolutionefficiency.Especiallyinthefaceofincreasinglycomplexengineeringproblems,manyexisting
4、algorithmswillalsobecomepowerless.Tosolvethisproblem,ahighefficientdynamicmicromulti-objectivegeneticalgorithmbasedontheMicroMulti-objectivegeneticalgorithmissuggested.Thealgorithmusessmall-scaleevolutionarypopulationtoimprovethesolutionefficiency,andtrackschanges
5、oftheenvironmentthroughanintroductionofanenvironmentdetectionmechanismintheevolutionaryprocesstoensurethatthealgorithmsearchnewParetooptimalsolutionsetquickly.Whenthepopulationofthecurrentgenerationaccomplishesthegeneticmanipulation,thedetectionmechanismwillbestar
6、tedforenvironmenttesting.Calculatethevalueofalltheindividualobjectivefunctionsandtheconstraintfunctions,andcomparetheresultswiththecorrespondingfunctionvaluecalculatedinthepreviousgeneration.Ifthereisanydifference,weconsidertheproblemhaschangedovertime.Atthistime,
7、selecttheappropriatewaytoregenerateapopulation,andcombinedwiththesavedinformationsolutionsofpreviousgenerationandexternalpopulations,updatenon.dominatedsolutionssetintoanewevolutionaryprocess.Then,fourdifferenttypesofdynamicmulti-objectiveoptimizationtestproblemsa
8、reusedtotesttheperformanceoftheefficientdynamicmicromulti.objectivegeneticalgorithm.AndbycomparisonwithDNSGA-II,anexistingdynamicmulti。objectiveoptimiza
此文档下载收益归作者所有