简介:Lowgainfeedbackreferstocertainfamiliesofstabilizingstatefeedbackgainsthatareparameterizedinascalarandgotozeroasthescalardecreasestozero.Lowgainfeedbackwasinitiallyproposedtoachievesemi-globalstabilizationoflinearsystemssubjecttoinputsaturation.Itwasthencombinedwithhighgainfeedbackindifferentwaysforsolvingvariouscontrolproblems.Theresultingfeedbacklawsarereferredtoaslow-and-highgainfeedback.Sincetheintroductionoflowgainfeedbackinthecontextofsemi-globalstabilizationoflinearsystemssubjecttoinputsaturation,therehasbeenefforttodevelopalternativemethodsforlowgaindesign,tocharacterizekeyfeaturesoflowgainfeedback,andtoexplorenewapplicationsofthelowgainandlow-and-highgainfeedback.Thispaperreviewsthedevelopmentsinlowgainandlow-and-highgainfeedbackdesigns.
简介:ThispaperinvestigatestherelationshipbetweenstatefeedbackandHamiltonicanrealization.Firest,itisprovedthatacompletelycontrollablelinearsystemalwayshasastatefeedbackstateequationHamiltonianrealization.NecessaryandsufficientconditionsareobtainedforittohaveaHamiltonianrealizationwithnaturaloutput.ThensomeconditionsforanaffinenonlinearsystemtohaveaHamiltonianrealizationaregiven.someconditionsforanaffinenonlinearsystemtohaveaHamiltonianrealizationaregiven.Forgeneralizedoutputs,theconditionsofthefeedback,keepingHamiltonian,arediscussed.Finally,theadmissiblefeedbackcontrolsforgeneralizedHamiltoniansystemsareconsidered.
简介:Inthispaper,supposingthatthereceivedsignalsattheinputareoftheformv(t)=s1[t,x1(t)]s2[t,x2(t)]+n(t)wheresiareFMsignals(i=1,2),anovelcross-coupledphase-lockedloop(CCPLL(M))anditsmathematicalmodelareobtained.Theglobalqualitativestructuralanalysisofthemathematicalmodelofthefirst-orderloop,theacqui-sitionregionandsynchronizationregionofthefirst-orderloop,andthesynchronizationregionofthesecond-orderloopareobtained.
简介:ThispaperisconcernedwiththeanalysisofafeedbackM[X]/G/1retrialqueuewithstartingfailuresandgeneralretrialtimes.Inabatch,eachindividualcustomerissubjecttoacontroladmissionpolicyuponarrival.Iftheserverisidle,oneofthecustomersadmittedtothesystemmaystartitsserviceandtherestjoinstheretrialgroup,whereasalltheadmittedcustomersgototheretrialgroupwhentheserverisunavailableuponarrival.Anarrivingcustomer(primaryorretrial)mustturn-ontheserver,whichtakesnegligibletime.Iftheserverisstartedsuccessfully(withacertainprobability),thecustomergetsserviceimmediately.Otherwise,therepairfortheservercommencesimmediatelyandthecustomermustleavefortheorbitandmakearetrialatalatertime.ItisassumedthatthecustomerswhofindtheserverunavailablearequeuedintheorbitinaccordancewithanFCFSdisciplineandonlythecustomerattheheadofthequeueisallowedforaccesstotheserver.TheMarkovchainunderlyingtheconsideredqueueingsystemisstudiedandthenecessaryandsufficientconditionforthesystemtobestableispresented.Explicitformulaeforthestationarydistributionandsomeperformancemeasuresofthesysteminsteady-stateareobtained.Finally,somenumericalexamplesarepresentedtoillustratetheinfluenceoftheparametersonseveralperformancecharacteristics.
简介:Thispaperaddressesanonlinearfeedbackcontrolproblemforthechaoticarchmicro-electro-mechanicalsystemwithunknownparameters,immeasurablestatesandpartialstate-constraintsubjectedtothedistributedelectrostaticactuation.Toreflectinherentpropertiesanddesigncontroller,thephasediagrams,bifurcationdiagramandPoincaresectionarepresentedtoinvestigatethenonlineardynamics.TheauthorsemployasymmetricbarrierLyapunovfunctiontopreventviolationofconstraintwhenthearchmicro-electro-mechanicalsystemfacessomelimits.AnRBFneuralnetworksystemintegratingwithanupdatelawisadoptedtoestimateunknownfunctionwitharbitrarilysmallerror.Toeliminatechaoticoscillation,aneuro-adaptivebacksteppingcontrolschemefusedwithanextendedstatetrackingdifferentiatorandanobserverisconstructedtolowerrequirementsonmeasuredstatesandprecisesystemmodel.Besides,introducinganextendedstatetrackingdifferentiatoravoidsrepeatedderivativeforthevirtualcontrolsignalassociatedwithconventionalbackstepping.Finally,simulationresultsarepresentedtoillustratefeasibilityoftheproposedscheme.
简介:Inthispaper,theauthorsdesignboundaryfeedbackcontrollersattheinteriornodetostabilizeastar-shapednetworkofEuler-Bernoullibeams.Thebeamsarepinnedeachother,thatis,thedisplacementsofthestructurearecontinuousbuttherotationsofthebeamsarenotcontinuous.Thewell-posed-nessoftheclosedloopsystemisprovedbythesemigrouptheory.Theauthorsshowthatthesystemisasymptoticallystableiftheauthorsimposeabendingmomentcontroloneachedge.Finally,theauthorsderivetheexponentialstabilityofthesystem.
简介:Thispaperstudiestheconsensusproblemofmulti-agentsystemsinwhichallagentsaremodeledbyagenerallinearsystem.Theauthorsconsiderthecasewhereonlytherelativeoutputfeedbackbetweentheneighboringagentscanbemeasured.Tosolvetheconsensusproblem,theauthorsfirstconstructastaticrelativeoutputfeedbackcontrolundersomemildconstraintsonthesystemmodel.Thentheauthorsuseanobserverbasedapproachtodesignadynamicrelativeoutputfeedbackcontrol.Iftheadjacentgraphofthesystemisundirectedandconnectedordirectedwithaspanningtree,withtheproposedcontrollaws,theconsensuscanbeachieved.Theauthorsnotethatwiththeobserverbasedapproach,someinformationexchangebetweentheagentsisneededunlesstheassociatedadjacentgraphiscompletelyconnected.