欢迎访问《兵工学报》官方网站,今天是

兵工学报 ›› 2001, Vol. 22 ›› Issue (3): 363-366.

• 论文 • 上一篇    下一篇

基于正交小波网络的非线性系统辨识新方法

宋执环、吕立华、李平   

  1. 浙江大学工业控制技术研究所,工业控制技术国家重点实验室,浙江杭州,310027
  • 收稿日期:2000-07-01 修回日期:2001-05-01 上线日期:2014-12-25
  • 通讯作者: 宋执环
  • 基金资助:
    国家自然科学青年基金资助项目(69704006)

A Novel Method for Identifying Nonlinear Systems Using Orthogonal Wavelet Networks

Song Zhihuan, Lu Lihua, Li Ping   

  1. National Key Laboratory of Industrial Control Technology, Institute of Industrial Process Control, Zhejiang University, Hangzhou, 310027
  • Received:2000-07-01 Revised:2001-05-01 Online:2014-12-25
  • Contact: Song Zhihuan

摘要: 在非线性系统辨识中,系统输入数据往往是非均匀分布的。对于此类问题,正交小波网络处理起来较复杂。本文基于小波网格系,提出两种近似方法,可简单有效地解决这一问题。该法最大特点是计算简单,便于应用。最后用于辨识非线性系统,仿真结果验证了该方法的可行性和有效性。

关键词: 小波网络 , 非均匀分布 , 小波多分辨分析 , 网格系

Abstract: In the identilication of nonlinear system, the system inputs are usually non-uniformly distributed. The approach using orthogonal wavelet networks to resolve the problem is complex. Tins paper presents two approximative methods based on wavelet grid structure to handle this problem sim?ply and effectively. The advantages of these methods are computationally efficient and easy to use. fi?nally ,nonlinear systems are identified. The results of simulation show that the provided methods are effective and feasible.

Key words: wavelet network , non-uniform distribution , wavelet multiresolution analysis , grid structure

中图分类号: