简介:TherehavebeenseveralmethodsfordeterminingTwave-end.ButnoneofthemcanovercomethedifficultyfrommultiformityofECGsignalpattern.Inthispaper,amethodfordeterminingTwave-endusingevolutionaryalgorithm(EA)isproposed.Inthisway,first,everycharacteristicparameterrelatedtoTwave-endisencodedtoastringofcodes,andadaptationfunctionisconstructedwiththestringofcodes.Thenchoosetheindividualaccordingtotheadaptationfunctionvalueanddogeneticoperation(reproduction,crossoverandmutation),soastoproduceoffspringwithmoreadaptationfunctionvalue.BecauseofEA'sautoadaptationandautoorganizationcharacter,itcantraceECGsignaltypeandfindtheTwave-endautomatically.ExperimentresultsshowthattheerrorratioofrecognizingTwave-endusingthismethodismuchsmallerthanthatusingexistingmethod.
简介:Comparedwithaccuratediagnosis,thesystem’sselfdiagnosingcapabilitycanbegreatlyincreasedthroughthet/kdiagnosisstrategyatmostkvertexestobemistakenlyidentifiedasfaultyunderthecomparisonmodel,wherekistypicallyasmallnumber.BasedonthePreparata,Metze,andChien(PMC)model,then-dimensionalhypercubenetworkisprovedtobet/kdiagnosable.Inthispaper,basedontheMaengandMalek(MM)?model,anovelt/k-faultdiagnosis(1k4)algorithmofndimensionalhypercube,calledt/k-MM?-DIAG,isproposedtoisolateallfaultyprocessorswithinthesetofnodes,amongwhichthenumberoffault-freenodesidentifiedwronglyasfaultyisatmostk.ThetimecomplexityinouralgorithmisonlyO(2nn2).
简介:UnlikethepreviousresearchworksanalyzingthestabilityoftheT-S(Takagi?Sugeno)fuzzymodel,anextensiononthestabilityconditionofT-Sfuzzysystemswithadifferentstrategyisprovided.Inthestrategyanewvariable,whichisrelativetothegradeoffuzzymembershipfunction,isintroducedtothestabilityanalysisandanewstabilityconclusionisdeduced.Thedefinitionofstabilityconditioninthispaperisdifferentfrompreviousworks,thoughtheyaresimilarinform.Withtheproposedmethod,thesimulationinflightcontrollawshowsabettereffectiveness.