摘要
提出了一种自校正型自适应控制法进行机器人机械手的位置和速度控制。复杂的机械手系统由一组离散时间的差分方程离散化,模型的参数由回归模型算法确定,而这种算法使模型误差的二次方的总和为最小。自适应控制器基于差分方程模型和选定的性能指标进行设计,控制器的增益使用系统参数的估计值在线计算。给出了PUMA560 机器人的计算机仿真结果,验证了此控制法的有效性。
Develops an automatic--correction type adaptive control tech-nique available to control a manipulator's position and velocity at each joint.The complicated manipulator system is modelled through discretizing a diffe-ronce equation in terms of a series of discrete time, with model's parametersdetermined by means of an on--line algorithm obtained from a regressive model--such an algorithm will minimize the sum of squared errors of the model.The adaptive controller is thus designed on a basis of the difference equationmodel with its performance criteria chosen properly. On--line computation isavailable to the controller's gains by estimating the system parameter values.The newly--developed control technique has been verified for its effectivenessby way of computer simulations conducted on a PUMA560 robot.
关键词
工业机器人
自适应控制
回归模型
industrial robot
adaptive control
regressive model
parameter estimate
computer simulation
