智能控制的机遇和挑战课件.ppt

《智能控制的机遇和挑战课件.ppt》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、TheOpportunitiesandChallengesOfIntelligentControlpresenter：吴大华MainContent1Introduction2DevelopmentofIntelligentControl5OpportunitiesandChallengesofIntelligentControl3ApplicationofIntelligentControl45ConclusionandOutlook1.1DefinitionofIntelligentControlW

2、hatIs"IntelligentControl"?IEEEDefinition:intelligentcontrolmusthaveasimulationabilityofhumanlearningandadaptability.Ingeneral,anintelligentcontrolsystemtohavefunctionsofsensitivitytotheenvironment,decision-makingandcontrol.Accordingtotheirperformancereq

3、uirementsofthedifferentlevels,wecanhaveavarietyofartificialintelligence.Analyzingandorganizingdataandtransformingdatatothestructureofinformationmachinecanunderstand;andthesystemcancontinuetoworkunderuncertainty.Identifyingobjectsandevents;acquiringandex

4、pressingknowledge;thinkingandplanfuturemovement.Abilityofperceivingtheenvironment,decision-makingandcontrolling.AdervancedHigherSimple1.2BackgroundofIntelligentControlClassicalControlTheoryModernControlTheoryIntelligentControlTheoryControlthedynamicsyst

5、emdescribedbydifferentialequationanddifferenceequation;Plantisthesingle-variablelinearsystemwithconstantcoefficients;Onlyappliedtosingleinputandsingleoutputcontrolsystem(SISO)PlantistheobjectchangedfromSISOtoMIMO;Receivingthesysteminformationwiththehelp

6、ofstatetransitionmodelnotthesensors;Methodschangedfromintegraltomatrixtheoryandgeometryandfromfrequencymethodtothemethodofstatespace;Transformingfrommechanismmodelingtostatisticalmodeling,startwithparameterestimationandsystemidentificationtheory;Forlarg

7、ecomplex,highdimensional,nonlinearplantandseriousuncertaintyplant.Independentonplantmodelandappliedfortheunknownandseriousuncertaintyobject;HasthecharacteristicsofhumanintelligenceCanexpressqualitativeknowledgeandhaveself-learningability;1.3Characterist

8、icsofIntelligentControlCharacteristicsofIntelligentControlCandoeffectiveglobalcontroltocomplexsystems(suchasnonlinear,fasttime-varying,multivariable,strongcoupling,uncertainty,etc),andhavebetterfaulttolerance;Multi-modelcontrolwi