机械科学与技术2026，Vol.45Issue(1)：20-26,7.DOI:10.13433/j.cnki.1003-8728.20240004
机械臂动力学前馈混合迭代学习控制研究
Study on Hybridization of Dynamic Feedforward with Iterative Learning Control for Manipulator
摘要
Abstract
In response to the problems in manipulator trajectory tracking control,including inaccurate dynamic parameters and unknown external disturbances,a control method integrating dynamic feedforward with iterative learning is proposed.Firstly,dynamic feedforward control was employed to enable the manipulator to achieve basic tracking of the desired trajectory,and the control system equations were linearized within a small vicinity of the desired trajectory.Subsequently,an iterative learning controller was introduced to mitigate the impact of inaccuracies in dynamic parameters and external disturbances.Through iterations,it further reduced the trajectory tracking errors.The error convergence of the present control scheme was analyzed by using a compression mapping approach.The simulation experiments were conducted by using M WORKS.Sysplorer platform,and the experiments were performed on a six-axis serial-link manipulator.The results indicated that the present approach,through iterative learning,gradually reduced the trajectory tracking errors of the manipulator.关键词
动力学前馈控制/迭代学习控制/机械臂/轨迹跟踪Key words
dynamic feedfoward control/iterative learning control/manipulator/trajectory tracking分类
信息技术与安全科学引用本文复制引用
严冬,陈立平,丁建完..机械臂动力学前馈混合迭代学习控制研究[J].机械科学与技术,2026,45(1):20-26,7.基金项目
国家重点研发计划(2019YFB1706501) （2019YFB1706501）
