摘要
本文将模糊算法和小脑模型神经网络有机地结合在一起 ,提出了一种单输入单输出 (SISO)的模糊小脑模型神经网络 (FCMAC) .它在对输入进行分级量化的同时进行模糊量化 ,利用 Takagi- Sugeno模糊算法进行推理 ,并将模糊算法引入 CMAC的权值训练 ,具有输入量化级数少、函数逼近精度高等特点 .这种FCMAC用于 Wiener模型辨识具有结构确定、计算量小、训练速度快、辩识效果好等特点 .
Combined with fuzzy control algorithm and Cerebellar Model Articulation Controller (CMAC), a kind of Fuzzy Cerebellar Model Articulation Controller (FCMAC) with Single Input Single Output (SISO) is proposed in this paper. It carries out input fuzzy quantification and grading quantification in the same time, deduces using Takagi Sugeno fuzzy algorithm, and introduces fuzzy algorithm into weight training of CMAC. It has the characteristics of less quantification grade and high accuracy of function approximation. It shows good effect by simulation used for Wiener model identifying, with fixed structure and quick training and little amount of calculation.
出处
《信息与控制》
CSCD
北大核心
2002年第2期159-163,共5页
Information and Control
