航空学报 > 1996, Vol. 17 Issue (5): 114-118

非线性系统的BP神经网络在线建模及故障检测

董选明, 裘丽华, 王占林   

  1. 北京航空航天大学303教研室, 北京, 100083
  • 收稿日期:1995-04-13 修回日期:1995-10-10 出版日期:1996-10-25 发布日期:1996-10-25

BP NEURAL NETWORKS BASED ON ON LINE MODELLING AND FAULT DETECTION IN NONLINEAR SYSTEMS

Dong Xuanming, Qiu Lihua, Wang Zhanlin   

  1. Faculty 303, Beijing University of Aeronautics and Astronautics, Beijing, 100083
  • Received:1995-04-13 Revised:1995-10-10 Online:1996-10-25 Published:1996-10-25

摘要: 回顾了非线性系统建模、故障检测与诊断(FDD)及BP神经网络应用中存在的问题。提出了一种BP网络在线阶段自学习方式及应用该方式对非线性系统进行在线建模和故障检测的方法。仿真结果表明,本文提出的方式和方法是可行的

关键词: 非线性系统, 神经网络, 故障检测

Abstract: The existing problems for modelling, fault detection and diagnosis, and BP neural networks′ application, in nonlinear systems have been reviewed. An on line stage self learning rule is presented, and a novel method for on line modelling and detecting faults using the above rule is also proposed. Simulations show that the presented rule and method are practical.

Key words: nonlinear systems neuralnet, faults-detection

中图分类号: