Chapter5Mechatronic Engineering
Chapter 5 Mechatronic Engineering
Contents:5.1Advanced Control andAutomation ofMechatronicSystems: 5.2 On-Line Monitoring and Fault DiagnosisofMechatronicSystems·5.3Fluid Transmission andControl:5.4 Embedded System and Intelligent Instrument
Contents • 5.1 Advanced Control and Automation of Mechatronic Systems • 5.2 On-Line Monitoring and Fault Diagnosis of Mechatronic Systems • 5.3 Fluid Transmission and Control • 5.4 Embedded System and Intelligent Instrument
5.1AdvancedControlandAutomationofMechatronicSystems:Maincontents.Terminology.Text Tour
5.1 Advanced Control and Automation of Mechatronic Systems • Main contents • Terminology • Text Tour
5.1AdvancedControlandAutomationofMechatronicSystemsMaincontents·5.1.1Control Systems·5.1.2PID Controller.5.1.3Brief IntroductionofAdvancedControl Methods
5.1 Advanced Control and Automation of Mechatronic Systems • Main contents • 5.1.1 Control Systems • 5.1.2 PID Controller • 5.1.3 Brief Introduction of Advanced Control Methods
PIDControllerKe(t)e(t)rAPlant/processKJee(t)drKdeodrFigure5.7AblockdiagramofaPIDcontrollerinafeedbackloopTheoverall controlfunctioncanbeexpressedmathematicallyasfollows:de(t)(5.1)u(t)=K,e(t)+K,f'e(t)dt+K,dtwhere Kp,K,and Kd,all non-negative,denote the coefficients forthe proportional,integralandderivativetermsrespectively(sometimesdenotedP,IandD)
PID Controller
ClassicAdvanced ControlMethodsadaptive control,fuzzylogic control,neural network control,etcReference modelW.(s)Plant PCommandInputcommandu(0)Strategy for adjustingPlantControllercontrollergainsyt)c(0)P(0)InputFigure5.11Adaptivecontrollerstructurewithadjustablecontrollergains7u(e)OnlineparameterestimationofgNBPBNMPMNS0(t)Calculations8(0-F(0(0)0-6-54-3Figure5.12IndirectMRACFigure5.13Eighttrianglefuzzysets
Classic Advanced Control Methods adaptive control, fuzzy logic control, neural network control, etc
MeasuredVialearningalgorithmLearningandsignals:controlInput,actionsoutput.andPerformancefeedbackerrorFigure5.14RepresentationoflearningandDendritesSynapsecontrolactionsinaneuralnetworkapproachNeuralSomainputsNonlinearTwo aspects ofaHopfieldnetworkareasx,(0)Threshold, weactivationw,functionfollows:1、SynapticweightsaredeterminedbyanalyticallyTosolving the constraints,not by performing an iterativeotherbeuronlearning process. The weights are fixed duringAxonal(neural)operationoftheHopfieldnetworkoutput2、Final network outputsare obtainedby operatingx,(t)Wfeedbacknetworks,forthesolutionofanapplicationtask.SynapticSomaticoperation(aggregation, thresholding,andoperationnonlinearactivation)
Two aspects of a Hopfield network are as follows: 1、Synaptic weights are determined by analytically solving the constraints, not by performing an iterative learning process. The weights are fixed during operation of the Hopfield network. 2、Final network outputs are obtained by operating feedback networks, for the solution of an application task
Commonly used neural network structuresInputOutputtofromotherotherneuronsorneuronsorsensorssensorsNeuralnodeSynapticweight(b)Astatic (feed-forward)neural network with n-inputs andm-outputs
Commonly used neural network structures
半技术词汇（Sub-technicalwords）--intergeral，半技术词汇是在各专业中出现频率都很高的词。这类词往往在不同的环境下有不同的含义。·intergeral1、名词：整体、积分integralcontrolP1212、形容词：必需的、不可或缺的；Control systems areanintegral partofmodern society作为组成部分的；完整的、完备的。积累plant
半技术词汇（Sub-technical words）- intergeral • 半技术词汇是在各专业中出现频率都很高的词。这类词往往在不同的环境 下有不同的含义。 • intergeral 1、名 词：整体、积分integral control P121 2、形容词：必需的、不可或缺的； Control systems are an integral part of modern society. 作为组成部分的； 完整的、完备的。 积累 plant
Terminology控制系统controlsystemskon'troul'sistams]controller[kan'troul]n.控制器someone who maintains and audits business accounts远程控制remotecontrol[rr'mautkan'trouln.工厂plant [plaint] buildings for carrying on industrial labor开环系统open-loopsystem['aupanlup'sistam]闭环系统closed-loopsystem[klauzd'lu:psistam]stability[sta'bilti]n.稳定性thequality or attribute of being/firm and steadfast瞬态响应transientresponse'traenziontrr'sppns]稳态误差steady-state error['stedi steit'ero()]功率放大poweramplificationl'pauaaemplaff'kerfon
Terminology