伺服系统的高斯基函数CMAC学习控制研究

On-Line Learning Control for Servo Systems based on CMAC with Gauss Basis Function

在高精度伺服系统中 ,由于摩擦力及负载扰动等因素的影响 ,常规PID控制难以满足越来越高的控制精度、跟踪性能等指标要求。本文提出了基于高斯基函数CMAC在线学习的控制方案 ,并给出了相应的学习算法。仿真结果表明 ,该方法不仅有较好的控制精度 。

The standard PID(Proportional-Integral-Derivative) type servo controller is not capable of delivering the desired precision and tracking performance under the influence of friction and load disturbance. In This paper, an on-line learning controller based on CMAC with Gauss Basis Function Neural Network is presented, and its algorithm is discussed. The simulation results show that this method not only can improve the precision efficiently, but also can use a high learning rate.