资源描述：
《Adaptive Control for a Class of Nonaffine Systems Based on Fuzzy-Neural Approach》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、IMACSMulticonferenceon"ComputationalEngineeringinSystemsApplications"(CESA),October4-6,2006,Beijing,China.AdaptiveControlforaClassofNonaffineSystemsBasedonFuzzy-NeuralApproachY.Q.Jin,J.H.Wu,W.J.GuDepartmentofControlEngineeringNavalAeronauticalEnginee
2、ringInstitute.YanTai,264001ChinaPhone:+86-535-6232638,Fax:+86-535-6232638,E-mail:jyq301gyahoo.com.cnAbstract-AnadaptivecontroldesignmethodisproposedReferasin[GE,02],adaptive-neuralcontrolmethodsforforaclassofuncertainsingle-inputsingle-output(SISO)ac
3、lassofpure-feedbacksystemswithgeneralnonaffinesystembasedonfuzzy-neuralapproach.Totheuncertaintieswereproposed.Butthesemethodshighlyauthors'knowledge,thiscontrolproblemisfirstlyrelyontheapproximationabilityofRBFneuralnetworksconsideredinthispaper.Iti
4、sconsidereddifficulttobedealtwithinthecontrolliterature,mainlybecausethatthevirtualanddon'ttakeuseofknowninformationefficiently.Atcontrolsandthefinalcontrollawofuncertainnonaffinethesametime,thesystemsconsideredin[GE,02]aresystemarenoteasytoresolve.T
5、oovercomethisdifficulty,affineincontrolu.thefuzzy-neuralapproximatorcancelstheunknownpartofInthispaper,wediscusstheadaptivedesignforaclassoftheinversefunctionsadaptively.Then,Inversedesign,uncertainSISOnonaffinesystembasedonfuzzy-neuralbacksteppingde
6、sign,andfeedbacklinearizationtechniquesapproach.Toauthor'sknowledge,noeffctivemethodforareincorporatedtodealwiththisproblem.Itisprovedthatthiscontrolproblemexistsintheliteratureatpresentthewholeclosed-loopsystemisstableinthesenseofstage.Thisismainlyd
7、uetothefactthatitisverydifficultLyapunov.Thecontrolperformanceisguaranteedbysuitablychoosingthedesignparameters.SimulationstudytofindvirtualcontrolsXidintermsofxiinwasincludedtodemonstratetheeffectivenessofthebacksteppingdesignprocedure.Toovercomethi
8、sproposedmethod.difficulty,fuzzy-neuralapproximatorisusedtoKeywords:Nonaffinesystem,fuzzy-neuralapproximator,approximatetheunknownpartofinversefunctions.Intheadaptivecontrol,backsteppingfinalstepn,feedbacklinearizationtechniquesandfuzzy-neuralapproxi
