简介:Introducingbasicdesignmethodologyfordevelopingbacksteppingnonlinearcontrollerforvibrationcontrolsystem.Withasimplifiedsecond-ordersystemmodelofnonlinearvibrationsystem(Duffing’sequation),where,theprocesshasillustratedthebacksteppingdesignstep-by-step.Backsteppingisanovelnonlineardesigntool,whichisbasedonconstructingtheLyapunovfunctionfortheclosed-loopsystemsandguaranteesthestabilityandtrackingperformancethroughenergydissipation.Ingeneral,thisnonlinearcontroldesignapproachgeneratesaggressivecontrolefforttoreducethetrackingerrorpresentedinthiscontrolsystemandsignificantlyimprovethesystembandwidth.Theeffectivenessofthedesignschemeisshownthroughthecomputersimulation.
简介:交换非线性的系统的一个类的稳定在纸被调查。担心的系统具有(概括)交换了Byrnes-Isidori正规形式,它都交换了模型在(概括)Byrnes-Isidori正规形式。首先,交换系统的稳定性结果被获得。然后,解决交换非线性的控制系统的稳定问题被用来。另外,必要;足够的条件为是等价于的反馈的一个交换仿射的非线性的系统被获得(概括)交换Byrnes-Isidori正规系统被介绍。最后,作为一个应用程序,交换lorenz系统的稳定性被调查。
简介:Acompoundneuralnetworkisutilizedtoidentifythedynamicnonlinearsystem.Thisnetworkiscomposedoftwoparts:oneisalinearneuralnetwork,andtheotherisarecurrentneuralnetwork.Basedontheinversetheoryacompoundinversecontrolmethodisproposed.Thecontrollerhasalsotwoparts:alinearcontrollerandanonlinearneuralnetworkcontroller.Thestabilityconditionoftheclosed-loopneuralnetwork-basedcompoundinversecontrolsystemisdemonstratedbasedontheLyapunovtheory.Simulationstudieshaveshownthatthisschemeissimpleandhasgoodcontrolaccuracyandrobustness.
简介:我们在场一个新歧管的学习算法叫了保存排列(LOPA)的本地Orthogonality。我们的算法被试图用仿射的转变排列多重本地邻居进一个全球坐标系统的本地正切空间排列(LTSA)方法启发。然而,LTSA经常没能保存象距离和角度那样的原来的几何数量。尽管为保存orthogonality的一个反复的排列过程被LTSA的作者建议,既不相应初始化也不实验被给。普罗克拉斯提斯Subspaces排列(PSA)实现由与退火模仿独立估计每旋转转变保存想法的orthogonality。然而,在PSA的优化是复杂、多重的分开的本地旋转可以生产全球性倾向於矛盾的结果。探讨这些困难,我们首先使用LTSA的伪逆诡计与统一全球坐标代表每本地直角的转变。第二,orthogonality限制被放松是半明确的编程(SDP)的一个例子。最后一个二拍子的圆舞反复的过程被采用进一步在直角的限制减少错误。广泛的实验证明LOPA能忠实地保存原来的数据集的距离,角度,内部产品,和邻居。在比较,当LOPA的运行时刻比PSA,MVU和MVE的显著地快时,LOPA的嵌入的表演比PSA的好、比得上象MVU和MVE一样的最先进的算法的。
简介:Non-intrusivemethodsforeyetrackingareimportantformanyapplicationsofvision-basedhumancomputerinteraction.However,duetothehighnonlinearityofeyemotion,howtoensuretherobustnessofexternalinterferenceandaccuracyofeyetrackingposetheprimaryobstacletotheintegrationofeyemovementsintotoday'sinterfaces.Inthispaper,wepresentastrongtrackingunscentedKalmanfilter(ST-UKF)algorithm,aimingtoovercomethedifficultyinnonlineareyetracking.IntheproposedST-UKF,theSuboptimalfadingfactorofstrongtrackingfilteringisintroducedtoimproverobustnessandaccuracyofeyetracking.ComparedwiththerelatedKalmanfilterforeyetracking,theproposedST-UKFhaspotentialadvantagesinrobustnessandtrackingaccuracy.Thelastexperimentalresultsshowthevalidityofourmethodforeyetrackingunderrealisticconditions.
简介:Focusislaidontheadaptivepracticaloutput-trackingproblemofaclassofnonlinearsystemswithhigh-orderlower-triangularstructureanduncontrollableunstablelinearization.Usingthemodifiedadaptiveadditionofapowerintegratortechniqueasabasictool,anewsmoothadaptivestatefeedbackcontrollerisdesigned.Thiscontrollercanensureallsignalsoftheclosed-loopsystemsaregloballyboundedandoutputtrackingerrorisarbitrarysmall.
简介:Inthispaper,nonlinearobserversareincorporatedintotheadaptivecontroltosynthesizecontrollersforaclassofuncertainnonlinearsystemswithunknownsinusoidaldisturbanceswhicharepresentedinmatchedandunmatchedforms.Inadditiontomagnitudesandphases,frequenciesofthesinusoidaldisturbancesneednotbeknownaswell,solongastheoverallorderisknown.Nonlinearobserversareconstructedtoeliminatetheeffectofunknownsinusoidaldisturbancestoimprovethesteady-stateoutputtrackingperformance-asymptoticoutputtrackingisachieved.Theadaptationlawisusedtoobtaintheestimateofallunknownparameters.Thepresenteddisturbancedecouplingalgorithmscandealwithmatchedandunmatchedunknownsinusoidaldisturbances.
简介:Adissipative-basedadaptiveneuralcontrolschemewasdevelopedforaclassofnonlinearuncertainsystemswithunknownnonlinearitiesthatmightnotbelinearlyparameterized.Themajoradvantageofthepresentworkwastorelaxtherequirementofmatchingcondition,I.e.,theunknownnonlinearitiesappearonthesameequationasthecontrolinputinastate-spacerepresentation,whichwasrequiredinmostoftheavailableneuralnetworkcontrollers.Bysynthesizingastate-feedbackneuralcontrollertonaketheclosed-loopsystemdissipativewithrespecttoaquadraticsupplyrate,thedevelopedcontrolschemeguaranteesthattheL2-gainofcontrolledsystemwaslessthanorequaltoaprescribedlevel.Andthen,itisshownthattheoutputtrackingerrorisuniformlyultimatebounded.Thedesignschemeisillustratedusinganumericalsimulation.
简介:在这篇论文,我们为冲动的控制系统以二项措施学习稳定性和围住的海角。由使用变化Lyapunov方法,稳定性和围住的海角上的一个新变化比较原则和一些标准被获得。一个例子被举说明建议结果的效率。
简介:为有未知结构的非线性的系统的一个系统鉴定方法用短输入产量数据被介绍。方法简化原来的NARMAX方法。它为非线性的系统介绍更一般的模型结构。方法被采用获得模型术语的数据处理的组方法(GMDH);参数。建议方法的有效性被一个典型非线性的系统与未知结构说明;缺乏的输入产量数据。