简介:Theincreasingusageofmultimediaobjects,suchasRSimages,scannedphotographs,videorecordings,textsandvoiceinformation,demandsanefficientmethodtoorganizeenormousinformation(Taylor,1996).Intwoyears,thesatelliteremotesensingindustrywillprovidegeographicinformationsystem(GIS)userswithasignificantnewtooltocreatebase_mapandupdatetheirGISdatabaseswithdetailedinformation.Thistoolwillbemoreandmoreusefulwhenthesatelliteimagerywith1mspatialresolutionisused(Kevin,1996). Hypermediatechniqueisfrequentlyusedtodealwithdifferentmedia,andtoorganizethediverseandinnumerabledataincomputerswithverylimitedCPUspeedandmemory.ThetechniquehasbeendemonstratedtobeanefficientwayforprocessingalargeamountofRSimagedata. Inthispaperwepresentthehypermediadatamodel,whichisbasedonRSimageinGIS.Usingthemodelconcept,wecanhandlealargeamountofRSimagedataefficientlywithverylimitedcomputingpower.Wehavedevelopedthemodelwiththreekindsofnodes:interiornode,physicalnodeandcomplexnodetoorganizetheinfinityimage.
简介:ThispaperconsidersH-infinitycontrolproblemforintervaldescriptorsystems.NecessaryandsufficientLMI-basedconditionsarederivedforquadratic-likeH-infinitycontrolanalysisofintervaldescriptorsystems.Usingtheanalysisresult,twotypesoffeedbackcontrollersaredesignedsothattheclosed-loopintervaldescriptorsystemsareadmissiblewithH-infinity-normlessthanaprescribedvalue.Tothebestoftheauthors’sknowledge,thisisthefirstpapertodealwiththeH-infinitycontrolproblemofintervaldescriptorsystemsintheliterature.
简介:这篇论文考虑源于易于致动器浸透和L无穷骚乱的线性系统的一个家庭的一个交换系统的骚乱忍耐/拒绝的问题。为骚乱的L无穷标准上的线性反馈获得,一个给定的切换的计划和给定的界限的一个给定的集合,条件被建立,以线性或双线性的矩阵不平等,在哪个下面一套某个形式面对致动器浸透和L无穷骚乱为一个给定的交换线性系统是不变的,并且靠近环的系统拥有骚乱拒绝的某个水平能力。与这些条件,反馈获得和切换的计划的设计能是抑制了优化问题被提出并且解决。骚乱忍耐被最大的界限在从一个给定的集合开始的轨道为仍然保持围住的骚乱上测量。骚乱拒绝被系统产量的L无穷标准或由驾驶它的状态进或在起源的一位小邻居以内保留它的系统鈥檚能力测量。结果那,在家庭的所有系统是相同的,交换系统还原剂到在一条切换的反馈法律下面的一个单个系统。模拟结果比一条单个线性反馈法律能证明在一条切换的反馈法律下面的如此的一个单个系统能有更强壮的骚乱忍耐/拒绝能力。关键词致动器浸透-骚乱拒绝-骚乱忍耐-L无穷性能-集合不变性-交换系统梁璐在沈阳出生了,中国,在1980。他从沈阳大学,沈阳,中国,和Huaqiao大学在自动控制收到了他的学士和硕士学位,Quanzhou,中国分别地。他当前在自动化,上海JiaoTong大学,和部门的部门向他的博士学位正在工作电并且计算机工程,弗吉尼亚大学。林宗莉是在部门的一个教授电并且在弗吉尼亚大学的计算机工程。他从厦门大学在数学和计算机科学收到了他的学士学位,中国厦门在1983,他在从空格技术的中国学院的自动控制的M.E学位,中国北京在1989,并且他的博士学位在电并且从华盛顿的计算机工程州立大学,华盛顿普尔曼,在1994。他的当前的研�
简介:WeproposedadynamicmodelidentificationanddesignofanH-Infinity(i.e.H_∞)controllerusingaLightweightPiezo-CompositeActuator(LIPCA).Asecond-orderdynamicmodelwasobtainedbyusinginputandoutputdata,andapplyinganidentificationalgorithm.TheidentifiedmodelcoincideswellwiththerealLIPCA.Toreducetheresonatingmodethatistypicalofpiezoelectricactuators,anotchfilterwasused.AfeedbackcontrollerusingtheH_∞controlschemewasdesignedbasedontheidentifieddynamicmodel;thus,theLIPCAcanbeeasilyusedasanactuatorforbiomemeticapplicationssuchasartificialmusclesormacro/micropositioninginbioengineering.Thecontrolalgorithmwasimplementedusingamicroprocessor,analogfilters,andpoweramplifyingdrivers.Oursimulationandexperimentalresultsdemonstratethattheproposedcontrolalgorithmworkswellinrealenvironment,providingrobustperformanceandstabilitywithuncertaindisturbances.
简介:Itisconcernedwiththeproblemofdisturbanceattenuationwithstabilityforuncertainnonlinearsystemsbyadaptiveoutputfeedback.Byapartial-stateobserverandBacksteppingtechnique,anadaptiveoutputfeedbackcontrollerwasconstructed,whichcansolvethestandardgaindisturbanceattenuationproblemwithinternalstability.
简介:这篇论文涉及在取样的大小下面在不明确的系统上过滤的柔韧的H无穷的问题,两连续骚乱;分离骚乱在系统被考虑。参数不确定性被假定变化时间围住标准。目的是设计一个asymptotically稳定的过滤器,用局部地取样的大小,它两个都保证柔韧的asymptotic稳定性;为为所有可被考虑的不确定性的过滤错误动力学的H无穷性能的规定水平。推导进程被介绍辅助系统简化;为如此的一个过滤器的存在的足够的条件被建议。在学习期间,主要结果被采用各种各样的矩阵技术表示为LMI。用Matlab软件的LMI工具箱,获得适当过滤器是很方便的。最后,一个数字例子证明方法是有效的;可行。
简介:ThispaperisconcernedwiththeH-infinitycontrolproblemforaclassofcascadeswitchednonlinearsystems.Eachswitchedsysteminthisclassiscomposedofazero-inputasymptoticallystablenonlinearpart,whichisalsoaswitchedsystem,andalinearizablepartwhichiscontrollable.ConditionsunderwhichtheH-infinitycontrolproblemissolvableunderarbitraryswitchinglawandundersomedesignedswitchinglawarederivedrespectively.Thenonlinearstatefeedbackandswitchinglawaredesigned.WeexploitthestructuralcharacteristicsoftheswitchednonlinearsystemstoconstructcommonLyapunovfunctionsforarbitraryswitchingandtofindasingleLyapunovfunctionfordesignedswitchinglaw.TheproposedmethodsdonotrelyonthesolutionsofHamilton-Jacobiinequalities.
简介:AnewapproachforrobustH-infinityfilteringforaclassofLipschitznonlinearsystemswithtime-varyinguncertaintiesbothinthelinearandnonlinearpartsofthesystemisproposedinanLMIframework.TheadmissibleLipschitzconstantofthesystemandthedisturbanceattenuationlevelaremaximizedsimultaneouslythroughconvexmulti-objectiveoptimization.TheresultingH-infinityfilterguaranteesasymptoticstabilityoftheestimationerrordynamicswithexponentialconvergenceandisrobustagainstnonlinearadditiveuncertaintyandtime-varyingparametricuncertainties.Explicitboundsonthenonlinearuncertaintyarederivedbasedonnorm-wiseandelement-wiserobustnessanalysis.
简介:H-infinityestimatorisgenerallyimplementedintimevariantstate-spacemodels,butitleadstohighcomplexitywhenthemodelisusedformultipleinputmultipleoutputwithorthogonalfrequencydivisionmultiplexing(MIMO-OFDM)systems.Thus,anH-infinityestimatorovertime-invariantsystemmodelsisproposed,whichmodifiestheKreinspaceaccordingly.InordertoavoidthelargematrixinversionandmultiplicationrequiredineachOFDMsymbolfromdifferenttransmitantennas,expectationmaximization(EM)isdevelopedtoreducethehighcomputationalload.JointestimationovermultipleOFDMsymbolsisusedtoresistthehighpilotoverheadgeneratedbytheincreasingnumberoftransmitantennas.Finally,theperformanceoftheproposedestimatorisenhancedviaanangle-domainprocess.Throughperformanceanalysisandsimulationexperiments,itisindicatedthattheproposedalgorithmhasabettermeansquareerror(MSE)andbiterrorrate(BER)performancethantheoptimalleastsquare(LS)estimator.JointestimationovermultipleOFDMsymbolscannotonlyreducethepilotoverheadbutalsopromotethechannelperformance.Whatismore,anobviousimprovementcanbeobtainedbyusingtheangle-domainfilter.
简介:ThispaperdiscussesH-infinitystatefeedbackcontrolforanetworkedcontrolsystemwithtime-varyingdelays.Basedonthefree-weighingmatrixmethod,adelay-dependentstabilitycriterionsatisfyingaprescribedH-infinitynormboundispresentedforanNCSwithunknown,time-varyingandboundeddelays.Andthen,thecriterionistransformedintosufficientconditionsbasedonlinearmatrixinequalitiesforH-infinitycontrol.TheconditionsthusobtainedarealsousedtodesignanH-infinitystatefeedbackcontroller.ThisdesignmethodisfurtherextendedtosolvethedesignproblemofrobustH-infinitystatefeedbackcontrol.Anumericalexampledemonstratesthevalidityofthemethod.
简介:ThispaperfocusesontherobustH-infinityreliablecontrolforaclassofnonlinearneutraldelaysystemswithuncertaintiesandactuatorfailures.Wedesignastatefeedbackcontrollerintermsoflinearmatrixinequality(LMI)suchthattheplantsatisfiesrobustH-infinityperformanceforalladmissibleuncertainties,andactuatorfailuresamongaprespecifiedsubsetofactuators.Anexampleisalsogiventoillustratetheeffectivenessoftheproposedapproach.