力矩受限的机器人吸引域估计方法  

Estimation methods of domain of attraction for robot with bounded torques

作　　者：彭文东[1] 苏剑波[1] 

机构地区：[1]上海交通大学自动化系,上海200240

出　　处：《控制与决策》2009年第7期979-984,共6页Control and Decision

基　　金：国家自然科学基金项目(60675041);教育部新世纪优秀人才支持计划项目(NCET-06-0398)

摘　　要：针对力矩受限的机器人组合非线性反馈控制的局部稳定区域描述问题,研究了吸引域的估计方法.利用不变集属性和椭球性质,定义两种不同意义的最大椭球不变集来逼近吸引域,分别采用设置初始状态法和参考形状集法求解.通过带有约束的优化问题描述,所有条,件均能转化为线性矩阵不等式条件,易于求解.由于采用优化技术,能够减小吸引域估计的保守性.数值算例验证了所提方法的有效性.To formulate the problem concerned with local stable region description of composite nonlinear feedback control for robot manipulators with bounded torques, the estimation methods of domain of attraction are studied. By using the properties of set invariance and ellipsoid, two different maximum ellipsoid invariant sets are defined as the estimation of domain of attraction, so the methods of setting initial state and reference shape set are introduced. Through the description for the constrained optimization problem, all conditions can be expressed in LMI conditions and conveniently solved. Owing to the utilization of optimization technique, both methods can reduce the conservatism of estimation. Simulation examples testify the effectiveness of the suggested methods.

关 键 词：机器人 有界力矩 组合非线性反馈 吸引域 线性矩阵不等式 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]