智能船舶路径跟踪自抗扰模型预测控制.pdfVIP

智能船舶路径跟踪自抗扰模型预测控制

摘要针对智能船舶在航行过程中为抵抗多源时变环境干扰与模型不确定性的影响，常引起控制力发生突变，致

使执行装置难以响应、船舶无法准确跟踪期望路径的问题，提出了一种自抗扰模型预测控制算法．该方法基于自

抗扰控制思想，设计修改型扩张状态观测器(MESO)对系统的状态和总未知扰动进行估计，并基于估计值设计鲁

棒补偿的模型预测控制算法(RC-MPC)．本文方法基于MESO，将复杂的船舶路径跟踪系统转化为含有干扰的线

性仿射系统；同时，为避免鲁棒模型预测控制导致结果保守，设计了观测误差鲁棒补偿算法，提高了控制器的干

扰抑制能力，增强了系统对模型失配的鲁棒性。结果证明了该自抗扰模型预测控制级联系统具有全局一致渐近稳

定性，仿真实验验证了算法的有效性。

关键词智能船舶；自抗扰；模型预测控制；欠驱动船舶；路径跟踪；级联系统稳定性；鲁棒补偿

Autodisturbancerejectionmodelpredictivecontrolforintelligentship

pathfollowing

AbstractDuringthevoyageoftheintelligentships,toresisttheinfluenceofmulti-sourcetime-varyingenvironmentalinterferences

andmodeluncertainties,thecontrolforcesareoftensuddenlychanged,whichmakesitdifficultfortheactuatorstorespond,andthe

shipcannotaccuratelyfollowthedesiredpath.Inresponsetotheaboveproblem,anautodisturbancerejectionmodelpredictive

controlalgorithmwasproposed.Thealgorithmwasbasedontheideaofactivedisturbancerejectioncontrol,amodifiedextended

stateobserver(MESO)wasdesignedtoestimatethestatesandthetotalunknowndisturbancesofthesystem,andarobust

compensationmodelpredictivecontrol(RC-MPC)algorithmwasdesignedbasedontheestimatedvalues.ByusingMESO,the

complexshippath-followingsystemistransformedintoalinearaffinesystemwithinterferences.Meanwhile,toavoidthe

conservativeresultscausedbytherobustmodelredictivecontrol,arobustcompensationalgorithmforobservationerrorswas

designed,whichimprovedthecontrollesinterferencesuppressioncapabilityandenhancedthesystemsrobustnesstomodel

mismatch.Resultsprovedthattheactivedisturbancerejectionmodelpredictivecontrolcascadesystemhadglobaluniform

asymptoticstability,andtheeffectivenessofthealgorithmwasverifiedbysimulationexperiments.

收稿日期2022-03-10.

56.