东北大学学报（自然科学版）

东北大学学报(自然科学版) ›› 2010, Vol. 31 ›› Issue (3): 305-308+316.DOI: -

• 论著 •    下一篇

带有稳定学习算法的小波神经网络及应用

丛秋梅;柴天佑;余文;   

  1. 东北大学信息科学与工程学院;东北大学流程工业综合自动化教育部重点实验室;墨西哥国立理工大学高级研究中心(CINVESTAV-IPN);
  • 收稿日期:2013-06-20 修回日期:2013-06-20 发布日期:2013-06-20
  • 通讯作者: -
  • 作者简介:-
  • 基金资助:
    国家重点基础研究发展计划项目(2009CB320601);;

Wavelet neural networks with stable learning algorithm and its application

Cong, Qiu-Mei (1); Chai, Tian-You (1); Yu, Wen (3)   

  1. (1) School of Information Science and Engineering, Northeastern University, Shenyang 110004, China; (2) Key Laboratory of Integrated Automation of Process Industry, Northeastern University, Shenyang 110004, China; (3) Departamento de Control Automatico, CINVESTAV-IPN, Mexico D F 07360, Mexico
  • Received:2013-06-20 Revised:2013-06-20 Published:2013-06-20
  • Contact: Cong, Q.-M.
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

935

摘要： 针对当系统存在未建模动态时,神经网络辨识易产生参数漂移和不稳定的问题,采用输入-状态稳定性(ISS,input-to-state stability)分析方法,获得小波神经网络权值矩阵和小波尺度参数的误差反传类时变学习算法,该算法不带有鲁棒修正即可以实现小波神经网络的鲁棒稳定性.仿真例子表明,此稳定学习算法优于一般的误差反传算法,并将带有稳定学习算法的小波神经网络用于污水处理过程出水水质COD(化学需氧量,chemical oxygen demand)的预测,获得了较好的效果.

关键词: 小波神经网络, 输入-状态稳定性, 稳定学习算法, 鲁棒稳定性, 污水处理过程, 化学需氧量

Abstract: In the presence of unmodeled dynamics, the parameters' drift even instability may occur in the identification system of neural networks. Input-to-state stability(ISS) approach is applied to achieving the error backpropagation-like time-varying learning algorithm of weight matrix and wavelet scaling parameters in wavelet neural networks, of which the robust stability is guaranteed without robust modification. Simulations showed that the stable learning algorithm outperforms conventional error backpropagation ones, and the application of the algorithm to the prediction of COD in wastewater treatment process gets good results.

中图分类号: