变负载机械手轨迹跟踪控制器设计

Trajectory tracking controller design for manipulator with variable loads

中文关键词  切换控制器  机械手  神经网络  平均驻留时间  自适应控制

英文关键词  switching controller  robotic manipulator  neural network  average dwell time  adaptive control

中文摘要

      本文针对具有变负载的不确定刚性机械手系统, 提出了一种依赖平均驻留时间的神经网络自适应切换控制策略. 本控制方案将夹持不同负载的刚性机械手系统视为切换系统, 即根据负载的不同将整个系统分为若干子系统, 并基于平均驻留时间原则对每个子系统分别设计控制器. 在各子系统中, 分别采用径向基函数(RBF)神经网络逼近系统结构参数, 以避免控制器对系统精确模型的依赖. 同时, 基于神经网络设计鲁棒补偿项, 以抑制集总扰动对系统的影响. 然后, 利用多Lyapunov函数方法证明了轨迹跟踪误差的一致最终有界性. 最后, 通过仿真验证, 所提出的控制方案不仅可实现变负载机械手期望轨迹的高精度跟踪, 而且可有效削弱输入力矩的抖振.

英文摘要

      In this paper, a neural network adaptive switching control strategy based on the average dwell time is proposed for uncertain rigid manipulator system with variable loads. In this proposal, the rigid manipulator system holding different loads is treated as a switched system, which is divided into several subsystems according to different loads, and then different sub controllers are designed for each subsystem based on the average dwell time principle. In every subsystem, the RBF neural network is utilized for approaching to the system structural parameters to avoid the dependence of the controller on accurate system model. Meanwhile, the RBF neural network is employed to design the robust compensation term to suppress the influence of lumped disturbance of the system. Then, the uniform final boundedness of trajectory tracking error is verified by multi-Lyapunov function method. Finally, the simulation results show that the proposed algorithm can not only achieve the high-precision tracking of the manipulator with variable loads, but effectively eliminate the chattering of input torque.