基于线性变参数H_∞反馈的机器人迭代学习控制器设计  

Iterative Learning Controller Design Based on LPV H_∞ Feedback for Robotic Manipulator

作　　者：虞忠伟[1] 陈辉堂[1] 王月娟[1] 

机构地区：[1]同济大学电子与信息工程学院,上海200092

出　　处：《西安交通大学学报》2001年第6期552-555,共4页Journal of Xi'an Jiaotong University

基　　金：国家自然科学基金资助项目 (6 96 85 0 0 4)

摘　　要：提出了一种基于线性变参数 (LPV)H∞ 反馈的迭代学习控制器 ,用于不确定性机器人的高精度轨迹跟踪 .此控制器包括反馈部分和前馈部分 ,其中反馈部分设计为LPVH∞ 控制 ,前馈部分设计为高阶PD型迭代学习控制 .在满足一定的收敛条件下 ,证明了该控制器的跟踪误差界正比于系统初始误差界和系统输出干扰项界 .仿真结果不仅验证了此控制器随机器人关节位置变化始终具有干扰衰减、鲁棒稳定的性能 ,而且还验证了此控制器具有高精度轨迹跟踪的性能 .An iterative learning controller based on gain scheduled linear parameter varying(LPV) H ∞ feedback for high precision trajectory tracking of uncertain robotic manipulator was presented.The controller included two parts.One was the feedback part which was designed to LPV H ∞control, the other was the feedforward part which was designed to high order PD type iterative learning control. This paper gave the convergence conditions of the proposed controller and proved that the tracking error bound was proportional to the bounds of reinitialization error and output distrubance. Simulation results demonstrate that the proposed controller possesses not only always disturb attenuation and robust stability performance as the joint positions vary, but also the high precision trajectory tracking performance.

关 键 词：机器人 线性变参数H∞反馈 迭代学习控制器 高阶PD型 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]