一种模糊小波神经网络辨识器的设计与应用  

The Design and Application of the Fuzzy Wavelet Neural Networks Identifier

作　　者：韩贵金[1] 周有[1] 

机构地区：[1]西安邮电学院信息与控制系,陕西西安710121

出　　处：《电子科技》2008年第7期50-52,共3页Electronic Science and Technology

摘　　要：针对现有模糊神经网络的语言变量层隶属函数的不足,借鉴小波变换良好的尺度变换和伸缩平移特性,提出了一种新的模糊小波神经网络模型,并研究了其结构和相应的优化算法。将该模糊小波神经网络作为系统辨识器进行计算机仿真,仿真结果表明所研究的模糊小波神经网络结构及其学习算法简单有效,系统辨识准确。Given the subject function deficiencies of the language variable layer in the existing fuzzy neural network, a new fuzzy wavelet neural network is put forward using the good dilation and translation characteristie of wavelet transformation with its structure and the corresponding optimization algorithm studied. The fuzzy wavelet neural network can be used as the identifier. Computer simulation results show that the structure and learning algorithm of the fuzzy wavelet neural network are simple and effective and the identification results are exact.

关 键 词：小波变换 模糊小波神经网络 辨识器 

分 类 号：TP332.3[自动化与计算机技术—计算机系统结构]