模糊神经元及其在FCMAC中的应用  

The Fuzzy Neuron and its Application in FCMAC

作　　者：赵承滨[1] 宋清昆[1] 

机构地区：[1]哈尔滨理工大学,黑龙江哈尔滨150080

出　　处：《自动化技术与应用》2003年第10期10-13,共4页Techniques of Automation and Applications

摘　　要：本文通过对模糊集合与神经网络的特点及相互关系的阐述 ,引出将二者优点结合于一身的模糊神经网络理论。从模糊神经元开始 ,着重介绍其拓扑结构、分类、隶属函数的特点及相应激励函数的确定方法。最后以模糊小脑神经网络 (FCMAC)为算例 ,阐述了模糊神经元的应用问题。仿真结果表明 :它具有比常规CMAC学习速度快 。By introducing the characteristics and relationship between the fuzzy sets and the neural network,this paper presents theory of fuzzy neural which combines their features.The topological structure of the fuzzy neuron its classification,the characteristic of membership function and the definition method of activation function are discussed.An example of FCMAC controller is given.The simulation result indicated that it is accurate and fast compared with the CMAC.

关 键 词：模糊神经元 FCMAC 模糊控制 模糊小脑神经网络 隶属函数 激励函数 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP273.4[自动化与计算机技术—控制科学与工程]