改进的粒子群优化控制算法及其仿真研究.pdf

《改进的粒子群优化控制算法及其仿真研究.pdf》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、机床与液压Oct．2012HydromechatronicsEngineeringVo1．40No．19DOI：10．3969／j．issn．1001—3881．2012．19．005ImprovedControlAlgorithmBasedonParticleSwarmOptimizationandItsSimulationResearchNXiankunPolytechnicSchool，ChongqingJiaotongUniversity，Chongqing400074，ChinaAbstract：Theperormanceofcon

2、trolsystemisdeterminedbythecontrolparameterofcontroler．Aimedatthepuzzleofparameterselectionforparticleswarmoptimization(PSO)controlalgo-rithmthatthephenomenonofprematureconvergencemadethebasicPSOalgorithmshavebeeneasytogetinlocaloptimalsolution，andresuRedinimpossibleconverg

3、encetoglobalextremum，aswelasbeingnotsohighinsearchprecisionandslowerinconvergencespeed，thepaperproposedasortofimprovedcontrolalgorithmbasedonparticleswarmoptimization．Inthepaper，itdiscussedtheparticleswarmoptimizationalgorithmswithgeneticthought(GAPSO)，researchedonimproveda

4、lgorithmofPSO，andmadethecomparativestudyforproposedcontrolalgorithmbymeansofsimulationexperiment．Theresponsecurveofsimulationresultdemonstratedthatitwouldbebetterincomparisonwithconventionalmethodindynamicandsteadyperformance，andverifiedthereasonabilityandfeasibilityoftheim

5、provedcontrolalgorithm．TheresearchresultshowsthattheimprovedalgorithmofPSOproposedbythepaperismoreefectiveforcontrollerparametertuning．Keywords：particleswarmoptimization，geneticthought，parametertuning，improvedcontrolal_gorithmofPSOthespeediness，accuracyandstability，thetechn

6、olo—1．Introductiongyofadvancedintelligentcontrolhasbeenobtainedwidegeneralizedapplication，andPSOalgorithmhasParticleswarmoptimization(PSO)[1]andthealwaysbeenusedincontrollerofintelligentcontrolgeneticalgorithm(CA)allaretheevolutionaryalgo-structuresoastomaketheparametertuni

7、ng[2]．rithmbasedonthetheoryofbiologyevolutionismandThegeneticalgorithmisasortofsearchmethodgeneticsetcforsolvingoptimizationproblem，andbasedonprinciplesofbiologyevolution，andithaseachofthemhasowncharacteristic．PSOalgorithmbetterabilityofglobaloptimizationandoptimizationhasl

8、otsofadvantagessuchasbeingfasterinconvey-strategyofrandomization[3-4]．Uptonow，GAal