简介:Abroadbandreflectivepolarizationconverterisproposed.Theunitcellofpolarizationconverteriscomposedofasingle-splitresonantring,adouble-splitresonantring,adielectricsubstrateandametallicground.Thesimulatedresultsshowthatthepolarizationconvertercanconvertx-polarizedwavesintoy-polarizedwavesandobtainabroadbandpolarizationconversionfrom0.5010THzto1.3900THzwiththepolarizationconversionratio(PCR)beyond80%atnormalincidence.Moreover,thesurfacecurrentdistributionsareinvestigatedtoexplainthepolarizationconversionmechanism.Finally,agoodagreementisachievedbetweensimulatedandmeasuredresults.Thepolarizationconvertercanbeappliedinterahertzimaging,communicationandstealthytechnology.
简介:Thispaperisconcernedwiththeglobalboundednessproblemforaclassofstochasticnonlinearsystemswithmatchedconditions.Theuncertaintiesinthesystemsareduetoparametervariationsandexternalstochasticdisturbance.Onlythematchedconditionsandthepossibleboundoftheuncertaintiesaredemanded.BasedonthestochasticLyapunovstabilitytheory,anexplicitcontrollerisconstructedinthegradientdirection,whichrendersresponsesoftheclosed-loopsystemsbegloballyboundedinprobability.Whenthesystemsdegradetolinearsystems,thecontrollerbecomeslinear.Illustrativeexamplesaregiventoshowtheeffectivenessoftheproposedmethod.
简介:Anovelopticaltestingtarget(OTT)withthreedegreesoffreedom(DOFs)isproposedformeasuringthetrackingperformanceofaphotoelectrictheodolite.Themaincomponents,suchastheazimuthaxissystem,thepitchingaxissystem,thelinearmotionsystem,andthesimulatedtargetgenerator,aredesigned.Furthermore,thelinearmodulewiththeformofalarge-spanandlow-stiffnesscantileverbeamisstrengthenedusinganintegratedoptimizationmethod.Afterthestructureisstrengthened,the1st-ordernaturalfrequencyincreasesfrom14Hzto32Hz.Finally,afunctionaltestisperformedandtheresultsshowthatcomparedwithatraditionalopticaltesttarget,thesimulatedtargetgeneratedbythenoveltargetisadvantageousinpractice.
简介:Arobustadaptivepredictorisproposedtosolvethetime-varyinganddelaycontrolproblemofanoverheadcranesystemwithastereo-visionservo.Thepredictorisbasedontheuseofarecurrentneuralnetwork(RNN)withtappeddelays,andisusedtosupplythereal-timesignaloftheswingangle.Therearetwotypesofdiscrete-timecontrollersunderinvestigation,i.e.,theproportional-integral-derivative(PID)controllerandtheslidingcontroller.Firstly,adesignprincipleoftheneuralpredictorisdevelopedtoguaranteetheconvergenceofitsswingangleestimation.Then,animprovedversionoftheparticleswarmoptimizationalgorithm,theparallelparticleswarmoptimization(PPSO)methodisusedtooptimizethecontrolparametersofthesetwotypesofcontrollers.Finally,ahomemadeoverheadcranesystemequippedwiththeKinectsensorforthevisualservoisusedtoverifytheproposedscheme.Experimentalresultsdemonstratetheeffectivenessoftheapproach,whichalsoshowtheparameterconvergenceinthepredictor.
简介:Inordertoassurequalityandcontrolprocessinthedevelopmentoftheaircraftcollaborativedesignsoftware,amaturityassessmentmodelisproposed.Therequirementsdesigning—houseofqualityisdesignedtoevaluatethematuritydegreeofthesolution,andtheevaluationresultscanhelptomanageandcontrolthedevelopmentprocess.Furthermore,afuzzyevaluationmethodbasedontheminimumdeviationisproposedtodealwiththefuzzyinformation.Thequantitativeevaluationresultofthematuritydegreecanbecalculatedbyoptimizingthesemanticdiscountfactoraimfortheminimumdeviation.Finally,thismodelisillustratedandanalyzedbyanexamplestudyoftheaircraftcollaborativedesignsoftware.
简介:Inordertodesignacomplexlaserresonatorwithmulti-parameters,themethodofparticleswarmoptimization(PSO)algorithmisemployed.Theparametersinfluencingtheresonatorstabilityandmodesizedistributionaretakenintoconsideration,andthestabilitycriteriaindexandthemodesizedistributionareusedastargetvalues.TheabsolutevaluesofthedifferencesbetweenpracticalandthetargetvaluesaresetasthefitnessfunctionforthePSO.Byminimizingthefitnessfunction,alaserresonatorwiththeoptimizedcavityparameterscanbefound.TheanalysesforthedesignexampledemonstratethefeasibilityandvalidityofthePSOmethodinthecomputeraideddesignofmul-ti-parameterslaserresonator.ApplyingPSOalgorithmintheintelligentdesignofsolidstatelaserresonatorscanrealizethe.transitionfrommanualtrial-and-errortocomputerintelligentdesignofthelaserresonators.
简介:新思科技日前宣布,Graphcore采用新思科技DesignPlatform成功设计其ColossusTM智能处理单元(IPU),与现有处理器(CPU和GPU)相比,加速了人工智能(AI)计算。新思科技全面的数字和定制设计流程通过关键的人工智能优化技术,使设计者能够在应用于云计算、数据中心、汽车和移动应用的AI芯片上,提供同类最佳的设计实现质量(QoR)和最短的设计收敛时间(TTR)。新思科技Fusion技术TM增强了AI芯片设计能力,包括互连规划、乘法累加(MAC)拓扑优化和完整的AIIP参考流程,以实现最快速度、最小面积、最低功耗及实现三者的最佳平衡。
简介:Arobustcontrollerforbanktoturn(BTT)missileswithaerodynamicfinsandreactionjetcontrolsystem(RCS)isdevelopedbasedonnonlinearcontroldynamicmodelscomprisingcouplingsandaerodynamicuncertainties.Thefixedtimeconvergencetheoryisincorporatedwiththeslidingmodecontroltechniquetoensurethatthesystemtracksthedesiredcommandwithinuniformboundedtimeunderdifferentinitialconditions.Unlikepreviousterminalslidingmodeapproaches,theboundofsettlingtimeisindependentoftheinitialstate,whichmeansperformancemetricslikeconvergenceratecanbepredictedbeforehand.Toreducetheburdenofcontroldesignintermsofrobustness,extendedstateobserver(ESO)isintroducedforuncertaintyestimationwiththeoutputsubstitutedintothecontrollerasfeedforwardcompensation.Cascadecontrolstructureisemployedwiththeproposedcontrollawandthereinthecompoundcontrolsignalisobtained.Afterwards,controlinputsfortwokindsofactuatorsareallocatedonthebasisoftheirinherentcharacteristics.Finally,anumberofsimulationsarecarriedoutanddemonstratetheeffectivenessofthedesignedcontroller.
简介:Acompositedintegratedguidanceandcontrol(IGC)algorithmispresentedtotackletheproblemoftheIGCdesigninthedivephaseforthebank-to-turn(BTT)vehiclewiththeinaccuracyinformationoftheline-of-sight(LOS)rate.Forthesakeoftheoreticalderivation,anIGCmodelinthepitchplaneisestablished.Thehigh-orderfinite-timestateobserver(FTSO),withtheLOSangleasthesingleinput,isemployedtoreconstructthestatesofthesystemonline.Besides,acompositedIGCalgorithmispresentedviathefusionofback-steppinganddynamicinverse.ComparedwiththetraditionalIGCalgorithm,theproposedcompositedIGCmethodcanattenuateeffectivelythedesignconservationoftheflightcontrolsystem,whiletheLOSrateismixedwithnoise.Extensiveexperimentshavebeenperformedtodemonstratethattheproposedapproachisgloballyfinite-timestableandstronglyrobustagainstparameteruncertainty.
简介:
简介:UnderType-IIprogressivelyhybridcensoring,thispaperdiscussesstatisticalinferenceandoptimaldesignonstepstresspartiallyacceleratedlifetestforhybridsysteminpresenceofmaskeddata.ItisassumedthatthelifetimeofthecomponentinhybridsystemsfollowsindependentandidenticalmodifiedWeibulldistributions.Themaximumlikelihoodestimations(MLEs)oftheunknownparameters,accelerationfactorandreliabilityindexesarederivedbyusingtheNewton-Raphsonalgorithm.Theasymptoticvariance-covariancematrixandtheapproximateconfidenceintervalsareobtainedbasedonnormalapproximationtotheasymptoticdistributionofMLEsofmodelparameters.Moreover,twobootstrapconfidenceintervalsareconstructedbyusingtheparametricbootstrapmethod.TheoptimaltimeofchangingstresslevelsisdeterminedunderD-optimalityandA-optimalitycriteria.Finally,theMonteCarlosimulationstudyiscarriedouttoillustratetheproposedprocedures.
简介:Aco-designschemeofevent-triggeredsamplingmechanismandactivefaulttolerantcontrol(FTC)isdeveloped.Firstly,afaultdiagnosisobserverisdesignedtoestimateboththefaultandthestatesimultaneouslybyusingtheevent-triggeredsampledoutput.SomeH∞constraintsbetweentheestimationerrorsandtheevent-triggeredsamplingmechanismareestablishedtoensuretheestimationaccuracy.Then,basedontheconstraintsandtheobtainedfaultinformation,anevent-triggereddetectorandastaticfaulttolerantcontrollerareco-designedtoguaranteethestabilityofthefaultysystemandtoreducethesensorcommunicationcost.Furthermore,theproblemoftheeventdetectoranddynamicFTCco-designisalsoinvestigated.Simulationresultsofanunstablebatchreactorarefinallyprovidedtoillustratetheeffectivenessoftheproposedmethod.