基于混合阈值事件触发机制的船舶自适应神经滑模控制.pdfVIP

基基基于于于混混混合合合阈阈阈值值值事事事件件件触触触发发发机机机制制制的的的船船船舶舶舶自自自适适适应应应神神神经经经滑滑滑模模模控控控制制制

摘要:针对存在时变干扰和通信受限的船舶艏向保持控制问题,本文设计了一种基于混合阈值事件触发机制的

船舶自适应神经滑模控制算法.首先,采用径向基函数神经网络(RBF-NNs)对控制系统的模型不确定部分进行在线

逼近.其次,构造了一种比例积分滑模面和基于混合阈值参数的事件触发规则来降低时变干扰引起的系统抖振和减

少传输信道的通信资源占用.并且在理论上分析了事件触发技术能够增强闭环控制系统对低频扰动的抗扰性.最

后,通过Lyapunov理论证明了所提控制算法能够保证所有误差信号满足半全局一致最终有界稳定(SGUUB).在时

变干扰条件下进行数值仿真,并且与现有控制技术进行对比,仿真结果表明,所提控制算法能够明显减少系统抖振

和控制输入频率.

关键词:欠驱动船舶;艏向保持;自适应控制;事件触发;滑模控制

AdaptiveneuralslidingmodecontrolforUSVwiththe

hybridthresholdevent-triggeredmechanism

Abstract:Thispaperinvestigatesanadaptiveneuralslidingmodecontrolalgorithmfortheheadingkeepingofthe

underactuatedsurfacevessels(USV)inpresenceoftheexternaldisturbanceandthecommunicationconstraint.Inthe

proposedalgorithm,thenonlineartermisapproximatedbyusingtheradialbasisfunctionneuralnetworks(RBF-NNs)

neuralnetworks.Furthermore,aproportionalintegralslidingmodetechniqueandhybridthresholdevent-triggeredrulehave

beenintroducedtoreducethesystemchatteringandchannelresourceoccupancy,whichcausedbytheexternaldisturbance

andsystemuncertainties.Inaddition,theprincipleofavoidinglowfrequencydisturbancethatusesaneventtriggered

mechanismisanalyzed.Finally,thesemi-globaluniformandultimatelybounded(SGUUB)stablehasbeenprovedby

employingtheLyapunovtheory.Thenumericalsimulationhasbeencarriedouttoverifytheeffectivenessoftheproposed

algorithm.Comparedwiththeexistingtechniques,itcanbenotedthattheproposedalgorithmiswiththeadvantagesin

aspectsofthesystemchatteringandthetransmissionfrequencyofthecontrolcommand.

Keywords:underactuatedvessel;headingkeeping;adaptivecontrol;event-triggeredcontrol;slidingmodecontrol