简介:Thequalityofskeletonsystemforthecheetahrobotgoeshandinhandwithitsbionicresultofitsshape,structureandfunctions.Inviewoftheskeletonsystemconstitutionandstructuralcharacteristicofthecheetah,theteamappliedstructuredesign,stimulationanalysisandparameteroptimizationtodevelopingthecheetahrobot.Inaddition,aftertheinventionofcheetahrobot'santeriorlumbarvertebrabasedonitsfunctionalattributeandconnectivityattribute,theSolidworksSimulationwasutilizedtoanalyzethedesign,accordingtowhichimprovementonthelumbarvertebrawasmade.Plus,theadvantagesoftheCADandCAEmadethehighefficiencyofdesignworkandhighqualityofthecheetahrobotpossible.
简介:TheMelnikovmethodwasextendedtoperturbedplanarnon-Hamiltonianintegrablesystemswithslowly-varyingangleparameters.Basedontheanalysisofthegeometricstructureofunperturbedsystems,theconditionoftransverselyhomoclinicintersectionwasestablished.ThegeneralizedMelnikovfunctionoftheperturbedsystemwaspresentedbyapplyingthetheoremonthedifferentiabilityofordinarydifferentialequationsolutionswithrespecttoparameters.ChaosmayoccurinthesystemifthegeneralizedMelnikovfunctionhassimplezeros.
简介:这份报纸论述新柔韧的滑动模式控制(SMC)有为一般不明确的变化时间的延期的开发得好的理论证明的方法有结构的无常和Brownian噪音(维纳进程)的随机的系统。建议方法的特色是把单个价值分解(SVD)用于所有结构的无常并且与SMC方法一起为控制设计介绍可调节的参数。它导致为为经由线性矩阵不平等(LMI)的一般不明确的随机的系统的柔韧的稳定性和一个新柔韧的控制器的状况形成的一个更少保守人士。系统状态能到达交换在概率1保证了的表面的SMC。而且,没有SVD,有SVD和可调节的参数的建议方法比方法是更少的保守主义,这理论上被证明。纸主要是为方法和结果提供所有严格的理论证明。
简介:Anewmethodforidentifyingnonlineartime-varyingsystemswithunknownstructureispresented.Themethodextendstheapplicationareaofbasissequenceidentification.Theessentialideaistoutilizethelearningandnonlinearapproximatingabilityofneuralnetworkstomodelthenon-linearityofthesystem,characterizetime-varyingdynamicsofthesystembythetime-varyingparametricvectorofthenetwork,thentheparametricvectorofthenetworkisapproximatedbyaweightedsumofknownbasissequences.Becauseofblack-boxmodelingabilityofneuralnetworks,thepresentedmethodcanidentifynonlineartime-varyingsystemswithunknownstructure.Inordertoimprovethereal-timecapabilityofthealgorithm,theneuralnetworkistrainedbyasimplefastlearningalgorithmbasedonlocalleastsquarespresentedbytheauthors.Theeffectivenessandtheperformanceofthemethodaredemonstratedbysomesimulationresults.
简介:Adiscretegain-varyingunknowninputobserver(UIO)methodispresentedforactuatorfaultdetectionandisolation(FDI)problemsinthispaper.Anovelresidualschemetogetherwithamovinghorizonthresholdisproposed.ThisdesignmethodologyisappliedtoanonlinearF16systemwithpolynomialaerodynamicscoefficientexpressions,wherethecoefficientexpressionsfortheF16systemandUIOsmaybeslightlydifferent.ThesimulationresultsillustratethatasatisfactoryFDIperformancecanbeachievedevenwhentheF16systemisundertheenvironmentofmodeluncertainties,exogenousnoiseandmeasurementerrors.
简介:Optimalcontrolsystemofstatespaceisaconservativesystem,whoseapproximatemethodshouldbesymplecticconservation.Basedonthepreciseintegrationmethod,analgorithmofsymplecticconservativeperturbationispresented.Itgivesauniformwaytosolvethelinearquadraticcontrol(LQcontrol)problemsforlineartime-varyingsystemsaccuratelyandefficiently,whosekeypointsaresolutionsofdifferentialRiccatiequation(DRE)withvariablecoefficientsandthestatefeedbackequation.Themethodissymplecticconservativeandhasagoodnumericalstabilityandhighprecision.Numericalexamplesdemonstratetheeffectivenessoftheproposedmethod.
简介:ThispaperinvestigatesaMultiple-InputMultiple-Output(MIMO)schemecombiningTransmitAntennaSelectionandreceiveMaximal-RatioCombining(TAS/MRC)intime-varyingRayleighfadingchannels.Wefirstpresentnewclosed-formexpressionsforoptimalreceivedSignal-to-NoiseRatio(SNR),whichisexpressedinpolynomialform.Theseareusedtoanalyzeergodiccapacity,outageprobabilityandBitErrorRate(BER)ofTAS/MRCsystems.Numericalresultsarepresentedtovalidatethetheoreticalanalysis.
简介:
简介:Inthispaper,thecharacteristicsofdifferentformsofmildslopeequationsfornon-linearwaveareanalyzed,andnewnon-lineartheoreticmodelsforwavepropagationarepresented,withnon-lineartermsaddedtothemildslopeequationsfornon-stationarylinearwavesanddissipativeeffectsconsidered.Numericalsimulationmodelsaredevelopedofnon-linearwavepropagationforwatersofmildlyvaryingtopographywithcomplicatedboundary,andtheeffectsarestudiedofdifferentnon-linearcorrectionsoncalculationresultsofextendedmildslopeequations.Systematicalnumericalsimulationtestsshowthatthepresentmodelscaneffectivelyreflectnon-lineareffects.
简介:ThispaperinvestigatestherobuststochasticstabilityandH∞analysisforstochasticsystemswithtime-varyingdelayandMarkovianjump.Byusingthefreeweightingmatrixtechnique,i.e.,He’stechnique,andastochasticLyapunov-Krasovskiifunctional,newdelay-dependentcriteriaintermsoflinearmatrixinequalitiesarederivedforthetherobuststochasticstabilityandtheH∞disturbanceattenuation.Threenumericalexamplesaregiven.Theresultsshowthattheproposedmethodisefficientandmuchlessconservativethantheexistingresultsintheliterature.更多还原
简介:Thefree-parameterperturbationmethodisappliedtosolvetheproblemsofnonlinearstabilityofsphericalshallowshellsunderuniformload.Asamodifiedperturbationmethod,thefree-parameterperturbationmethodenablesresearcherstoobtainallcharacteristicrelationswithoutchoosingthecertainperturbationparameter.Someexampleswerediscussedtostudythevarietyregulationsofdeflectionsandstressofshellsintheprocessofbuckling,andtheresultswerecomparedwiththoseofotherresearchers.
简介:APlasmaSurfaceInteraction(PSI)experimentonChinaLowActivationMarten-sitic(CLAM)steelwasdonetocheckifCLAMsteelcouldbeusedasaPlasmaFacingMaterial(PFM).Aspecimenwithadiameterof45mmwasexposedto897shotsofdeuteriumplasmaswithatotaldurationof712secataminorradiusof30cminHT-7tokamak.Duringtheexpo-sureexperiment,noobservableinfluencewasfoundonplasmaperformance.Afterexposure,thesurfaceofthespecimenseemedassmoothasbeforebutwithsomecolourchangeatthemarginofthespecimen.Eventhoughsomemicro-damage,suchasdenseblisters,melting,splashing,depositions,anddust,wasfoundonlocalsurfaceswithScanningElectronMicroscopic(SEM)observation.Thereflectivityofthespecimendecreasedonlyslightly.AlloftheseshowsCLAMsteelhasgoodstabilityandirradiationresistance.Withfurtheroptimization,itcouldpossiblybeusedasthefirstmirrormaterialforplasmadiagnosticsintokamaks.
简介:Evolutionarycomputationbasedontheideaofbiologicevolutionisonetypeofglobaloptimizationalgorithmthatusesself-adaptation,self-organizationandrandomsearchingtosolveoptimizationproblems.Theevolutionary-simplexalgorithmisintroducedinthispaper.Itcontainsfloatingencodingwhichcombinestheevolutionarycomputationandthesimplexalgorithmtoovcrcomctheproblemsencounteredinthegeneticalgorithmandevolutionarystrategymethods.Numericalcxpcrimcntsarcperformedusingseventypicalfunctionstoverifythealgorithm.Aninverseanalysismethodtoidentifystructuralphysicalparametersbasedonincompletedynamicresponsesobtainedfromtheanalysisinthetimedomainisprcscntcdbyusingtheevolutionary-simplexalgorithm.Themodalevolutionary-simplexalgorithmconvertedfromthetimedomaintothemodaldomainisproposedtoimprovetheinverseefficiency.Numericalcalculationsfora50-DOFsystemshowthatwhcncomparedwithothermethods,theevolutionary-simplexalgorithmoffersadvantagesofhighprecision,cfficientsearchingability,strongabilitytoresistnoise,independenceofinitialvalue,andgoodadaptationtoincompleteinformationconditions.
简介:同样质量增加了的为系统的要求,对的需要高,系统可靠性也正在增加。软件系统是极其重要的,以提高的可靠性和稳定性,为提供高质量为顾客的服务。因为软件系统的复杂性,然而,软件开发能费时间、昂贵。许多统计模型在过去的年里被开发了估计软件可靠性。在这份报纸,我们与操作环境的无常建议一个新三参数的差错察觉软件可靠性模型。建议模型的明确的吝啬的值函数解决方案被介绍。例子被介绍给基于三说明建议模型和几个存在非同类的泊松进程(NHPP)模型的goodness-of-fit从软件应用程序收集的失败数据设定。建议模型适合的结果表演显著地好与其它存在NHPP基于三当模特儿比标准象均方差(MSE)那样,预兆的比率风险(PRR),和预兆的力量(PP)。