简介:Bothnecessaryandsufficientmaximumprinciplesforoptimalcontrolofstochasticsystemwithrandomjumpsconsistingofforwardandbackwardstatevariablesareproved.Thecontrolvariableisallowedtoenterbothdiffusionandjumpcoefficients.Theresultisappliedtoamean-varianceportfolioselectionmixedwitharecursiveutilityfunctionaloptimizationproblem.Explicitexpressionoftheoptimalportfolioselectionstrategyisobtainedinthestatefeedbackform.
简介:Marginalriskrepresentstheriskcontributionofanindividualassettotheriskoftheentireportfolio.Inthispaper,weinvestigatetheportfolioselectionproblemwithdirectmarginalriskcontrolinalinearconicprogrammingframework.'Theoptimizationmodelinvolvedisanonconvexquadraticallyconstrainedquadraticprogramming(QCQP)problem.WefirsttransformtheQCQPproblemintoalinearconicprogrammingproblem,andthenapproximatetheproblembysemidefiniteprogramming(SDP)relaxationproblemsoversomesubrectangles.InordertoimprovethelowerboundsobtainedfromtheSDPrelaxationproblems,linearandquadraticpolarcutsareintroducedfordesigningabranch-and-cutalgorithm,thatmayyieldane-optimalglobalsolution(withrespecttofeasibilityandoptimality)inafinitenumberofiterations.ByexploringthespecialstructureoftheSDPrelaxationproblems,anadaptivebranch-and-cutruleisemployedtospeedupthecomputation.Theproposedalgorithmistestedandcomparedwithaknownmethodintheliteratureforportfolioselectionproblemswithhundredsofassetsandtensofmarginalriskcontrolconstraints.
简介:Theproblemofconstructingamodelreferenceadaptivecontrollawforanuncertain1-dimensionalparabolicsystemisconsideredinthisarticle.Thecontrollerdesignedhereinvolvesonlytheplantstatebutnoitsderivatives.Aprioriboundsontheplant'suncertainparametersareusedtoproposeswitchinglawswhichserveasanadaptivemechanism.Theexponentialdecaytozeroofthestateerrorwithanyprescribedrateisguaranteedbychoosingacontrollerparametercorrespondingly.Numericalstudiesarealsopresentedtoillustratetheapplicabilityofthecontrollaw.
简介:这份报纸涉及提交向后的问题推迟系统包含的最佳的控制推动控制。作者为这种系统建立一个随机的最大的原则。建议问题的大多数区分的特征是控制变量由常规、冲动的控制组成,与时间延期,并且常规控制的域未必是凸的。作者为最佳的控制获得必要、足够的条件,它在数学金融有潜在的应用。
简介:Thispaperinvestigatestheleader-followingconsensusproblemformulti-agentsystemswithevent-triggeredmechanismandcontrolpacketlosses.Basedonthelocalsynchronizationerror,event-triggeredmechanismsareproposedinordertoreducethenumberofcontrollerupdate.Thecontrolpacketmayloseduetounreliabilityofcommunicationchannel.Withtheassumptionthatoncethepacketlosshappensthecontrollerwillbesettozero,sufficientconsensuscriteriaformulti-agentsystemwithevent-triggeredmechanismandcontrolpacketlossesisobtained.Itisalsoshownthattheinterplayamongtheallowablepacketlossrate,event-triggeredmechanismandsamplingperiod.Anillustrativeexampleisgiventodemonstratetheeffectivenessofthetheoreticalresults.
简介:CRITICALBIRTHORDERRATIOCONTROLANDATTAINABLESETOFADISCRETEAGE-DEPENDENTPARITYPSOGRESSIONPOPULATIONSYSTEMGUOBaozhu(Departmentof...
简介:与non-holonomic限制为联网的平面多代理人系统追踪控制问题的时间变体一致在这份报纸被调查。在追踪问题的时间变体一致,一个领导人代理人被期望同时,追踪需要的引用输入追随者代理人被期望维持时间变体形成。解决与non-holonomic限制追踪平面多代理人系统的问题的时间变体一致,追踪控制策略的时间变体一致根据单向性的拓扑学结构被设计。这份报纸的主要贡献之一是与non-holonomic限制为平面多代理人系统的一般时间变体形成追踪协议的时间变体一致,这份报纸的另外的主要贡献是活跃预兆的控制策略,在代理人的预言活跃地被产生的地方,以便计算效率比被动途径被改进。建议控制策略被wheeled的变化时间的形成的二种类型验证活动机器人,和试验性的结果证明建议控制策略为在本地、世界范围的联网的环境与non-holonomic限制追踪平面多代理人系统的问题的一般时间变体一致是有效的。
简介:Thispaperaddressestheproblemonsensorfaultestimationandfault-tolerantcontrolforaclassofTakagi-SugenoMarkovianjumpsystems,whicharesubjectedtosensorfaultsandpartiallyunknowntransitionrates.First,theoriginalplantisextendedtoadescriptorsystem,wheretheoriginalstatesandthesensorfaultsareassembledintothenewstatevector.Then,anovelreduced-orderobserverisdesignedfortheextendedsystemtosimultaneouslyestimatetheimmeasurablestatesandsensorfaults.Second,byusingtheestimatedstatesobtainedfromthedesignedobserver,astate-feedbackfault-tolerantcontrolstrategyisdevelopedtomaketheresultingclosed-loopcontrolsystemstochasticallystable.Basedonlinearmatrixinequalitytechnique,algorithmsarepresentedtocomputetheobservergainsandcontrolgains.Theeffectivenessoftheproposedobserverandcontrollerarevalidatedbyanumericalexampleandacomparedstudy,respectively,andthesimulationresultsrevealthattheproposedmethodcansuccessfullyestimatethesensorfaultsandguaranteethestochasticstabilityoftheresultingclosed-loopsystem.