航空学报 > 2005, Vol. 26 Issue (2): 195-198

基于神经网络和专家系统的电传操纵系统故障诊断

徐荣红, 孙金标   

  1. 空军指挥学院 战术教研室, 北京 100089
  • 收稿日期:2004-02-05 修回日期:2004-10-15 出版日期:2005-04-25 发布日期:2005-04-25

Fault Diagnosis for the Fly-by-Wire System Based on Artificial Neural Network and Expert System

XU Rong-hong, SUN Jin-biao   

  1. Faculty of Tactics, Air Force Command College, Beijing 100089, China
  • Received:2004-02-05 Revised:2004-10-15 Online:2005-04-25 Published:2005-04-25

摘要: 提出了一种神经网络与专家系统相结合的故障诊断方法,研制了相应的诊断模块,并在某型飞机电传操纵系统的故障诊断中得以实现。在实物仿真和实际诊断实例中,该诊断模块均达到了很好的效果。从而证明了所提出的方法作为一种诊断策略,可作为飞机电传操纵系统快速故障诊断的基础。该故障诊断方法可推广应用到其它同类飞机的故障诊断当中,为现代航空维修保障的故障诊断提供了一条有效的新途径。

关键词: 非定常气动力, 拓扑结构图, Hopf分叉, 旋转阻尼

Abstract: Based on the artificial neural network and expert system theories, a new fault diagnosis method for the fly-by-wire system is put forward and applied to the development of a fault diagnosis module for a operational aircraft with fly-by-wire flight control system. The module can be used to find out the fault parts of the fly-by-wire system correctly and quickly in the physical simulations and practical diagnosis cases, and hence, the provided idea and approach are proved to be suitable for fault diagnosis. Moreover, this fault diagnosis method can also be applied to other similar aircraft and serves as a new effective method for the fault diagnosis of modern aircraft.

Key words: fly-by-wire, fault diagnosis, neural network, expert system

中图分类号: