简介:Thispaperstudiestheexponentialstabilityofintervaltime-varyingdynamicalsystemwithmultidelay.Bythematrixmeasureanddelaydifferentialinequality,somesufficientconditionsforexponentialstabilityofintervaltime-varyingdynamicalsystemwithmultidelayareestablished.Theseconditionsareanimprovementandextensionoftheresultsachievedinearlierpapers.Finally,anumericalexampleisgiventodemonstrateourresult.
简介:Thecontrollabilityandobservabilityofnetworkedcontrolsystemsarestudied.Aimingatthenetworkedcontrolsystemwithtime-varyingdelay,thesufficientandnecessaryconditionsforcompletecontrollabilityandcompleteobservabilityofthesystemarepresented,respectively.BecauseofMarkovcharacteristicofthenetwork-induceddelay,intermsofstochastictheory,asufficientandnecessaryconditionforcompletelymeanvaluecontrollabilityofnetworkedcontrolsystemsisobtained.Further,theconditionsthatthecontrollabilityandobservabilityofnetworkedcontrolsystemsareequivalenttotheinitialtime-invariantsystemaregiven.Controllabilityandobservabilityrealizationindexesarealsodiscussed,respectively.Thenumericalexampledemonstratestheeffectivenessoftheproposedtheory.
简介:Newdelay-independentanddehy-dependentstabilitycriteriaforlinearsystemswithmultipletime-varyingdelaysareestablishedbyusingthetime-domainmethod.Theresultsarederivedbasedonanew-typestabilitytheoremforgeneralretardeddynamicalsystemsandnewanalysistechniquesdevelopedintheauthor'spreviouswork.Unlikesomeresultsintheliterature,alloftheestablishedresultsdonotdependonthederivativeoftime-varyingdelays.Therefore,theyaresuitableforthecasewithveryfasttime-varyingdelays.Inaddition,someremarksarealsogiventoexphintheobtainedresultsandtopointoutthelimitationsofthepreviousresultsintheliterature.
简介:Inthispaper,exponentialstabilityofHopfield-typeneuralnetworkswithtime-varyingdelaysareanalyzed.ByusingtheLyapunovfunctionalmethod,sufficientconditionsareobtainedforgeneralexponentialstabilities.Atthesametime,theoutputfunctionsdonotsatisfytheLipschitzconditionsanddonotrequiretherntobedifferentialorstrictlymonotonouslyincreasing.Moreover,allresultsareestablishedwithoutassuminganysymmetryoftheconnectionmatrix.Amtmericexampleispressentedtoshowtheeffectiveofthesecriteria.
简介:Radio-frequency(RF)tomographyisanemergingtechnologywhichderivestargetslocationinformationbyanalyzingthechangesofreceivedsignalstrength(RSS)inwirelesslinks.ThispaperpresentsandevaluatesanovelRFtomographysystemwhichiscapableofdetectingandtrackingatime-varyingnumberoftargetsinaclutteredindoorenvironment.ThesystemincorporatesanobservationmodelbasedonRSSattenuationhistogramandamulti-targettracking-by-detectionfilteringapproachbasedonprobabilityhypothesisdensity(PHD)filter.Inaddition,thesequentialMonteCarlomethodisappliedtoimplementthemulti-targetfiltering.Toevaluatethetrackingsystem,theexperimentsinvolvingupto3targetswereperformedwithinanobstructedindoorareaof70m^2.Theexperimentalresultsindicatethattheproposedtrackingsystemiscapableoftrackingatime-varyingnumberoftargets.
简介:Consideringthejointchannelestimationanddatadetectionintime-varyingorthogonalfrequencydivisionmultiplexing(OFDM)andaddressingtransmissionperformancedegradationinducedbythesevereinter-carrierinterference(ICI)atveryhighspeed,anewprogressiveiterativechannelestimationschemeisproposed.ToalleviatetheerrorpropagationoftheinaccuratedataduetoICI,themeasurementsubcarriersintheKalmanfilterisdesignedtobeextendedfrompilotssubcarrierstoallthesubcarriersprogressivelythroughtheiterations.Furthermore,initerationprocess,theinterferenceofthenon-pilotdatatothemeasurementsubcarriersisconsideredtobepartofnoiseinthemodifiedKalmanfilter,whichimprovestheestimationaccuracy.Simulationindicatesthattheproposedschemeimprovestheperformanceinfasttime-varyingsituation.
简介:
简介:Inthepapertheproblemontheassignmentoftheboundsofdecreasingrateforatime-varyinglinearcontrolsystemisdiscussed.Thesufficientandnecessaryconditionforboundsofdecreasingrateofatime-varyinglinearsystemtobeassignedarbitrarilyispresented.Itispointedoutthatforanygivenrealnumbern,M,n
简介:Thispaperstudiesscale-typestabilityforneuralnetworkswithunboundedtime-varyingdelaysandLipschitzcontinuousactivationfunctions.Severalsufficientconditionsfortheglobalexponentialstabilityandglobalasymptoticstabilityofsuchneuralnetworksontimescalesarederived.Thenewresultscanextendtheexistingrelevantstabilityresultsinthepreviousliteraturestocoversomegeneralneuralnetworks.
简介:Thispaperisconcernedwiththeadaptivestabilizationproblemofuncertaininputdelayedsystems.Asolutiontothisproblemisgivenforaclassofuncertainnonlinearsystemswithtime-varyingdelaysinbothstateandinput.Anadaptiveasymptoticallystabilizingcontroller,whichcanguaranteethestabilityoftheclosed-loopsystemandtheconvergenceoftheoriginalsystemstate,isdesignedbymeansoftheLyapunov-Krasovskiifunctionalstabilitytheorycombinedwithlinearmatrixinequalities(LMIs)an...
简介:Theproblemofdelay-dependentrobuststabilityforsystemswithtitne-varyingdelayhasbeenconsidered.ByusingtheS-procedureandthePark'sinequalityintherecentissue,adelay-dependentrobuststabilitycriterionwhichislessconservativethanthepreviousresultshasbeenderivedfortime-delaysystemswithtime-varyingstructureduncertainties.Thesameideahasalsobeeneasilyextendedtothesystemswithnonlinearperturbations.Numericalexamplesillustratedtheeffectivenessandtheimprovementoftheproposedapproach.
简介:Inthispaper,westudythecontrollabilityresultsforthenonlinearimpulsiveintegrodifferentialevolutionsystemswithtime-varyingdelaysinBanachspaces.Thesufficientconditionsofexactcontrollabilityisprovedunderwithoutassumingthecompactnessoftheevolutionoperator.TheresultsareobtainedbyusingthesemigrouptheoryandtheSchaferfixedpointtheorem.
简介:Thispaperisconcernedwiththedesignofamemorystatefeedbackcontrollerforlinearsystemswithintervaltime-varyingdelays.Thetimedelayisassumedtobeatime-varyingcontinuousfunctionbelongingtoagiveninterval,whichmeansthatthelowerandupperboundsoftime-varyingdelayareavailable.First,alessconservativedelay-range-dependentstabilitycriteriaisproposedbyusinganewintervalfractionmethod.Intheprocessofcontrollersynthesis,thehistoryinformationofsystemisconsideredinthecontrollerdesignbyintroducingthelowerdelaystate.Moreover,theusualmemorylessstatefeedbackcontrollerfortheunderlyingsystemscouldbeconsideredasaspecialcaseofthememorycase.Finally,twonumericalexamplesaregiventoshowtheeffectivenessoftheproposedmethod.