个人简介
席在荣，男，1969年9月8日出生，现为中国科学院数学与系统科学研究院研究员。
1988年湖南常德师范学校毕业，
1991年湖南省高等教育自学考试数学专科毕业(国防科技大学主考)，
1993年湖南省高等教育自学考试数学本科毕业(国防科技大学主考)，
1997年7月于郑州大学获得理学硕士学位，
2000年7月于中国科学院数学与系统科学研究院获得控制论专业博士学位，
2002年5月在清华大学电气工程博士后流动站完成博士后工作，
2004年获教育部自然科学一等奖，
2005年获教育部自然科学二等奖及山东省自然科学三等奖，
2008年获国家自然科学二等奖等，
2006-2008年曾任IEEEControlSystemSociety,BeijingChapter秘书长，
IEEESeniorMember;
中国自动化学会控制理论专业委员会委员;
中国人工智能学会智能空天系统专业委员会委员;
《系统科学与数学》、《控制理论和应用》编委等学术职务。
主持并参加过国家自然科学基金、国家自然科学基金重大项目、重点项目及973项目等，
已发表学术论文100余篇，包括IEEETransactionsonAutomaticControl，Automatica，IEEETransactionsonControlSystemsTechnology，JournalofPhysicsAMathematical&Theoretical及PhysicalReviewA等国际著名学术刊物。
研究领域
非线性系统分析和合成
电力系统和机械系统
量子控制
量子机器学习
近期论文
查看导师新发文章
（温馨提示：请注意重名现象，建议点开原文通过作者单位确认）
Passivity-basedstabilizationandH8controloftheHamiltoniancontrolsystemswithdissipationanditsapplicationstopowersystems
ThispaperconsidersthestabilizationandH1controloftheHamiltoniancontrolsystemswithdissipation.Certain
NonlineardecentralizedcontrollerdesignformultimachinepowersystemsusingHamiltonianfunctionmethod
Inthispaper,wefirstexpressamultimachinepowersystemasaHamiltoniancontrolsystemwithdissipation.Then,usingtheHamiltonianfunctionmethodadecentralizedexcitationcontrolscheme,asastaticmeasurablefeedback,isproposedtostabilizethemultimachinepowersystem.Then,itisshownthatthecontrolschemewithproperlychosenparametersisalsoanH∞control,whichsolvestheproblemofdisturbanceattenuationsimultaneously.Finally,thedesigntechniqueisdemonstratedbyathree-machinepowersystem.
Aswitchingalgorithmforglobalexponentialstabilizationofuncertainchainedsystems
Thisnotedealswithchainedformsystemswithstronglynonlinearunmodeleddynamicsandexternaldisturbances.Theobjectiveistodesignarobustnonlinearstatefeedbacklawsuchthattheclosed-loopsystemisgloballyKexponentiallystable.Weproposeanovelswitchingcontrolstrategyinvolvingtheuseofinput/statescalingandintegratorbackstepping.ThenewfeaturesofourcontrollersincludetheabilitytoachieveLyapunovstability,exponentialconvergence,androbustnesstoasetofuncertaindriftterms.
Non-Markovianentanglementdynamicsbetweentwocoupledqubitsinthesameenvironment
Weanalyzethedynamicsoftheentanglementintwoindependentnon-Markovianchannels.Inparticular,wefocusontheentanglementdynamicsasafunctionoftheinitialstatesandthechannelparameters,suchasthetemperatureandtheratiorbetweenω0,thecharacteristicfrequencyofthequantumsystemofinterest,andωc,thecut-offfrequencyoftheOhmicreservoir.Wegiveastationaryanalysisoftheconcurrenceandfindthatthedynamicofnon-Markovianentanglementconcurrence{\calC}_{\rho}(t)attemperaturekBT=0isdifferentfromthekBT>0case.Wefindthat'entanglementsuddendeath'(ESD)dependsontheinitialstatewhenkBT=0,otherwisetheconcurrencealwaysdisappearsatafinitetimewhenkBT>0,whichmeansthattheESDmusthappen.Themainresultofthispaperisthatthenon-MarkovianentanglementdynamicisfundamentallydifferentfromtheMarkovianone.IntheMarkovianchannel,entanglementdecaysexponentiallyandvanishesonlyasymptotically,butinthenon-Markovianchanneltheconcurrence{\calC}_{\rho}(t)oscillates,especiallyinthehightemperaturecase.Thenanopen-loopcontrolleradjustedbythetemperatureisproposedtocontroltheentanglementandprolongtheESDtime.
Theentanglementdynamicsofthebipartitequantumsystem:towardentanglementsuddendeath
Weinvestigatetheentanglementdynamicsofthebipartitequantumsystembetweentwoqubitswiththedissipativeenvironment.WebeginwiththestandardMarkovianmasterequationintheLindbladformandtheinitialstatewhichispreparedintheextendedWerner-likestate:ρΦAB(0).Weexaminetheconditionsforentanglementsuddendeath(ESD)andcalculatethecorrespondingESDtimebytheWoottersconcurrence.WeobservethatESDisdeterminedbytheparameterssuchasthemeanoccupationnumberoftheenvironmentN,amountofinitialentanglementαandthepurityr.ForN=0,wegettheanalyticalexpressionsofbothESDconditionandESDtime.ForN>0wegiveatheoreticalanalysisthatESDalwaysoccurs,andsimulatetheconcurrenceasafunctionofγ0tandoneoftheparametersN,αandr.
OptimalDecoherenceControlinnon-MarkovianOpen,DissipativeQuantumSystems
Weinvestigatetheoptimalcontrolproblemfornon-Markovianopen,dissipativequantumsystem.
Nonlineardecentralizedsaturatedcontrollerdesignforpowersystems
Adirectmethodofconstructingadecentralizedsaturatedcontrollerisveryimportantinthecontrollerdesignofmultimachinepowersystems.Amongtheexistingdesignstheauthorfirstconstructedastatefeedback,thensaturatedthecontroller.However,theauthorcouldnotprovethatthissaturatedcontrollerwasvalid.Inthispaperamulti-machinepowersystemasaHamiltoniancontrolsystemwithdissipationisfirstrepresented.ThenusingaHamiltonianfunctionadecentralizedsaturatedexcitationcontrolscheme,whichisastaticmeasurablefeedback,isproposed.Last,thedesignisdiscussedindetailonathree-machinepowersystem.Simulationsshowthatthesaturated,simplecontrollawworkswell.
Onoutputfeedbackstabilizationofuncertainchainedsystems."InControl,Automation
Thispaperdealswithchainedformsystemswithstronglynonlineardisturbancesanddriftterms.Theobjectiveistodesignrobustnonlinearoutputfeedbacklawssuchthattheclosed-loopsystemsaregloballyexponentiallystable.Thesystematicstrategycombinestheinput-state-scalingtechniquewiththeso-calledbacksteppingprocedure.
Globaladaptiveoutputregulationofaclassofnonlinearsystemswithnonlinearexosystems
Thispaperdealswithglobaloutputregulationwithnonlinearexosystemsforaclassofuncertainnonlinearoutputfeedbacksystems.Thecirclecriterionisexploitedfortheinternalmodeldesigntoaccommodatethenonlinearitiesintheexosystems,andtheexplicitconditionsaregivenfortheexosystemssuchthattheproposedinternalmodeldesigncanbeapplied.Theuncertaintiesoftheoutputfeedbacksystemsareintheformofunknownconstantparameters,andadaptivecontroltechniquesareusedtoensuretheglobalstabilityoftheproposedcontroldesignforoutputregulation.
Outputfeedbackexponentialstabilizationofuncertainchainedsystems
Thispaperdealswithchainedformsystemswithstronglynonlineardisturbancesanddriftterms.Theobjectiveistodesignrobustnonlinearoutputfeedbacklawssuchthattheclosed-loopsystemsaregloballyexponentiallystable.Thesystematicstrategycombinestheinput-state-scalingtechniquewiththeso-calledbacksteppingprocedure.Adynamicoutputfeedbackcontrollerforgeneralcaseofuncertainchainedsystemisdevelopedwithafilterofobservergain.Furthermore,twospecialcasesareconsideredwhichdonotusetheobservergainfilter.Inparticular,aswitchingcontrolstrategyisemployedtogetaroundthesmoothstabilizationissue(difficulty)associatedwithnonholonomicsystemswhentheinitialstateofsystemisknown.
Classicalandquantumcontrolofasimplequantumsystem
Aqubitwhichispreparedinoneoftwonon-orthogonalstatesandsubjectedtobit-flippingnoiseisconsidered.Theobjectiveistousemeasurementandfeedbackcontroltocorrectthestateofthequbit.Threeclassicalschemesusingprojectivemeasurements,i.e.discriminationandre-preparation,donothingandrandompreparation,havebeendiscussed,andarenotoptimalwithrespecttoaperformancewhichisquantifiedbytheaveragefidelityofthecorrectedstatecomparedtotheinitialstate.Inaddition,onequantumschemeusinganon-projectivemeasurementwithanoptimummeasurementstrengthachievesthebesttrade-offbetweengaininginformationaboutthesystemanddisturbingitthroughmeasurementback-action.Theperformanceofaquantumcontrolschemeoutperformstheclassicalschemes.Furthermore,nouniversalcorrectedschemeisdiscussed.
Adaptiverejectionofnon-lineardisturbancesinextendednon-linearoutputfeedbacksystems
Thispaperdealswithasymptoticrejectionofdisturbancesgeneratedfromnon-linearexosystemsforuncertainnonlinearsystemsinanextendedoutputfeedbackform,whichallowsthevectorfieldcoupledwiththesysteminputtohavedifferentnonlinearfunctionsofthesystemoutputasitselements.Anewinternalmodeldesignisproposedtodealwithnonlinearfunctionsofthesystemoutputthatarecoupledwiththeinputandtheunknowndisturbance.Adaptivecontroltechniquesarethenusedtodealwiththeuncertaintyinthesystem.Theproposedadaptivedisturbancerejectionalgorithmwiththenewinternalmodeldesignensurestheasymptoticrejectionoftheunknownnonlineardisturbanceandtheboundednessofallthevariables.
Controlledpopulationtransferforquantumcomputinginnon-Markoviannoiseenvironment
Weproposeacontrolapproachtotransferthepopulationbetweenselectedquantumstatesinthenon-Markovianopenquantumsystem.Thistransfers,assistedbysinglequbitphaseshiftoperationscangenerateuniversallogicgatesforquantumcomputing.Wefindthattheoccupationprobabilitybehavesdifferentlyfordifferentenvironmentalconditions,suchasthetemperatureandtheratiobetweenthefrequencyofthequantumsystemofinterest,andthecut-offfrequencyofOhmicreservoir.Weprovideanactivecontrolovernon-Markovianopenquantumsystemforthequantumstateengineering.
OptimalDynamicalDecouplingSequenceforOhmicSpectrum
Weinvestigatetheoptimaldynamicaldecouplingsequenceforaqubitcoupledtoanohmicenvironment.
Dynamicalcontrolofquantumcorrelationsinacommonenvironment
京公网安备 11010802027423号