基于Transputer的CMAC神经网络模型的高速高精度直接驱动机器人的运动控制研究

APPLICATION OF CMAC NEURAL NETWOKS BASED ON TRANSPUTERS UTILIZED IN THE D.D ROBOT CONTROL

本文主要研究了利用智能学习方法（ＣＭＡＣ神经网络）获得高品质运动控制的问题，重点解决了ＣＭＡＣ的快速实现问题；用Ｔｒａｎｓｐｕｔｅｒ和ＣＭＡＣ网络构成并行数字控制器，对两关节的直接驱动机器人进行了实验研究，结果表明，ＣＭＡＣ网络学习控制方法是一种有效的和实用的高速高精度运动控制方法，并具有良好的鲁棒性。

In this paper, we discussed a motion controller with high performance by using intelligent learning method(CMAC neural network). The problesm of quick-implementing of CMAC network is solved here. The CMAC networks based on robotic digital parallel controller using Transputer developped are successfully used in high speed moving control of two-joint direct-drive robot.System experimental results demonstrate that CMAC neural network is an effective and robust scheme in high speed and accurate robotic moving control.