近空间飞行器泛函连接网络自适应预测控制_英文_

《近空间飞行器泛函连接网络自适应预测控制_英文_》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、Jun.2010TransactionsofNanjingUniversityofAeronautics&AstronauticsVol.27No.2ADAPTIVEPREDICTIVECONTROLOFNEAR-SPACEVEHICLEUSINGFUNCTIONALLINKNETWORKDuYanli,WuQingxian,JiangChangsheng(CollegeofAutomationEngineering,NUAA,29YudaoStreet,Nanjing,210016,P.R.China)Abstract:Anovelnonlinearadaptive

2、controlmethodispresentedforanear-spacehypersonicvehicle(NHV)inthepresenceofstronguncertaintiesanddisturbances.Thecontrollawconsistsoftheoptimalgeneralizedpredic-tivecontroller(OGPC)andthefunctionallinknetwork(FLN)directadaptivelaw.OGPCisacontinuous-timenonlinearpredictivecontrollaw.TheF

3、LNadaptivelawisusedtooffsettheunknownuncertaintiesanddistur-bancesinaflightthroughtheonlinelearning.Thelearningprocessdoesnotneedanyofflinetrainingphase.ThestabilityanalysesoftheNHVclose-loopsystemareprovidedanditisprovedthatthesystemerrorandtheweightlearningerrorareuniformlyultimatelyb

4、ounded.Simulationresultsshowthesatisfactoryperformanceofthecon-trollerfortheattitudetracking.Keywords:predictivecontrolsystems;adaptivecontrolsystems;uncertainty;functionallinknetwork;near-spacevehicleCLCnumber:V249.1Documentcode:AArticleID:1005-1120(2010)02-0148-07INTRODUCTIONcontrolpr

5、oblemsofNHVs.Recently,variousgeneralizedpredictivecon-Theflightcontrollerdesignofnear-spacehy-trol(GPC)algorithmshavebeendeveloped.personicvehicles(NHVs)hasbecomeoneoftheTheirmajorshortcomingisrequiringtheonlinegreatestaeronauticalresearchchallenges.Amongdynamicoptimization.Ref.[4]prese

6、ntedakindthenear-space(20—100kmabovetheearth),ofclosed-formoptimalGPC(OGPC)lawtoavoidtheresistanceuponNHVisrelativelylow,onlinecomputation,thusbeingeffectiveforfastwhereasthereexistrarefiedair,complicatedtem-time-varyingobjects,especiallyfortheattitudeperaturevariationandunexpecteddistu

7、rbances.systemsofNHVs.Nevertheless,uncertaintiesMoreover,thelackofabroadflightdynamicwerenotconsideredinRef.[4].Inthispaper,databaseaddssignificantparameteruncertaintiesOGPCandanewfunctionallinknetwork(FLN)totheNHVmodel.Someinvestigatorsproposedadaptivecontrollawformarobustpred