摘要
本文讨论分解运动速度控制框架下将梯度投影法应用于冗余机器人控制的有关问题．实施梯度投影法的关键之一是自运动大小的选择．本文先分析了确定自运动大小的现有方法的优缺点，然后提出按机器人关节配置奇异程度来确定自运动优化因子的思想和方案．这种新方案与原有的动态优化方案相比，更有利于快速、并行的实施，文中给出了相应的计算思路．计算机仿真比较了该方案与固定优化因子方案的控制结果．
Application of gradient projection method to control of redundant robots is discussed under theframework of resolved motion rate control (RMRC). One of key issues in applications of gradient projectionmethod is determination of self-motion. In the paper,existing methods to determine self-motion are analyzedfirst. Then a new idea and relative scheme are proposed to decide self-motion optimization factor according tojoint configuration of robots. This scheme has advantages over exist ones because it is more appropriate to implement gradient projection method fast and parallel. Such a parallel formulation is presented. Numerical simulation shows the efficiency of the scheme.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
1998年第2期184-189,共6页
Control Theory & Applications
关键词
梯度投影法
冗余机器人
速度控制
运动控制
gradient projection method
RMRC
redundancy resolution
parallel computation
singular value decomposition
