【摘要】 基于ＣＭＡＣ，设计了一种神经网络控制器，对常用的ＣＭＡＣ权值学习方法进行修改，提出了一种基于目标函数优化的控制器权值算法。当被控对象的特性未知时，用神经网络辨识器进行辨识。并且给出了整个控制系统的自适应步骤。仿真实验表明了方案的有效性。更多还原

【Abstract】 Based on Cerebellar Model Articulation Controller (CMAC), a kind of design scheme of neural network controller is given. The traditional method of modifying the weights of CMAC is improved according to the minimization of a cost function. When the characteristics of the object to be controlled is unknown, it can be identified by a neural network idetifier. The self-adaptive algorithm of the whole control system is proposed. The simulation experiment demonstrates the effectiveness of this scheme. 更多还原