一种CMAC网络模型及其在机器人控制中的应用

AN CMAC NETWORK MODEL AND ITS APPLICATION TO ROBOTIC CONTROL

小脑模型关节控制器(CMAC)具有学习算法简单、在线学习速度快的优点，非常适于机器人等复杂系 统的自适应控制．本文阐述了 CMAC 的原理，证明了其收敛性，提出了一种适合于机器人轨迹跟踪控制的 CMAC，并给出了仿真实验结果．

The cerebellar model articulation oontroller （CMAC） has simple learning algorithm and high on- line learning speed. It is very useful to the .self- adaptive oontrol of oomplicated plant such as robot. In this paper, the principle of CMAC was described,and the leaning oonvergence was proved. An CMAC designed for robotic manipulator tracking oontrol was proposed,and the simulations results were presented.