基于运动速度的机器人学习控制

为了使机器人跟踪给定的期望轨线，提出了一种新的基于机器人运动重复性的学习控制法．在这种方法中机器人通过重复试验得到期望运动，这种控制法的优点：一是对于在期望运动附近非线性机器人动力学的近似表达式的线性时变机械系统产生期望运动的输入力矩可不由估计机器人动力学的物理参数形成；二是可以适当的选择位置、速度和加速度反馈增益矩阵，从而加快误差收敛速度；三是加入了加速度反馈，减少了速度反馈，减少了重复试验的次数．这是因为在每次试验的初始时刻不存在位置和速度误差，但存在加速度误差．另外，这种控制法的有效性通过ＰＵＭＡ５６２机器人的前三个关节的计算机仿真结果得到验证。

To make a robot track a given desired motion trajectory, a new learning control scheme is proposed which is based on the repeatability of robot motion. In this scheme the tobot obtains a desired motion by repeating trials. This control method has three advantages: First, this control scheme is that the input torque that generates the desired motion can be formed without estimating the physical parameters of a linear time-varying mechanical system, which is an approximate representation of nonlinear robot dynamics in the vicinity of the desired motion ;Second, the rate of error convergence may be guaranteed rapidly by setting appropriate angular position, velocity and acceleration feedback gain Matrices;Third, in this control method, acceleration feedback is edded to decrease the number of repeated trials. This is duo to the fact that no position and velocity error exist at the beginning of every trial movement. Moreover, effectiveness of this control scheme is demonstrated through computer simulation in which a PUMA 562 manipulator with first three degrees of freedom is used.

中国科学院机器人学开放研究实验室资助