首页 | 本学科首页   官方微博 | 高级检索  
     

小波网络的研究进展与应用
引用本文:刘志刚,王晓茹,钱清泉.小波网络的研究进展与应用[J].电力系统自动化,2003,27(6):73-79,85.
作者姓名:刘志刚  王晓茹  钱清泉
作者单位:西南交通大学电气化自动化研究所,四川省成都市,610031
基金项目:国家自然科学基金资助项目 (5 99770 19)
摘    要:首先从小波分析和神经网络各自存在的问题出发,对小波网络的产生原因和产生形式进行了研究;接着分别从连续小波变换,正交小波变换,小波框架和小波基拟合几方面,详细地介绍了小波网络的构造理论;然后从小波函数的选择,网络参数初始化,隐层节点数确定和参数调节算法几方面对小波网络的学习过程进行了讨论;最后介绍了小波网络的应用,并提出了当前存在的问题和今后的研究方向。

关 键 词:小波网络  小波变换  函数逼近  学习算法
收稿时间:1/1/1900 12:00:00 AM
修稿时间:1/1/1900 12:00:00 AM

A REVIEW OF WAVELET NETWORKS AND THEIR APPLICATIONS
Liu Zhigang,Wang Xiaoru,Qian Qingquan.A REVIEW OF WAVELET NETWORKS AND THEIR APPLICATIONS[J].Automation of Electric Power Systems,2003,27(6):73-79,85.
Authors:Liu Zhigang  Wang Xiaoru  Qian Qingquan
Abstract:First, the arisen reasons and forms of wavelet networks are investigated based on faults of wavelet analysis and neural network in this paper. Then, the construction theories of wavelet networks are introduced in detail from continuous wavelet transformation, orthogonal wavelet transformation, wavelet frame and wavelet bases fitting. And the learning processes of wavelet networks are discussed, including choice of wavelet functions, initialization of networks parameters, choice of hidden layer node number and regulating algorithm of networks parameters. In the end, the applications of wavelet networks are presented, and the existent problems and future study aspects are proposed. This project is supported by National Natural Science Foundation of China (No. 59977019).
Keywords:wavelet network  wavelet transformation  function approximation  learning algorithm
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《电力系统自动化》浏览原始摘要信息
点击此处可从《电力系统自动化》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号