一种神经网络专家系统的故障诊断方法

The Diagnostic Expert System Based on Artificial Neural Network

提出将设备故障分为底层结构故障和高层功能故障的新概念，在此基础上研究用人工神经网络（ＡＮＮ） 诊断设备结构故障和用专家系统（ＥＳ） 诊断设备功能故障的设备故障诊断新方法，同时讨论了故障特征信号的获取和故障知识的组织，并以ＭＫ９ － ５ 卷烟机为对象进行实验研究，取得了较好的效果。

Anew ideais proposedfor devicefault diagnosis,the devicefaults were dividedinto thefunction alfaults and constructionalfaults.Based on it,a novelapproach ofintelligent diagnosis wasinvestigated .The di agnostic expertsystem based on Artificial Neural Networks(ANN) is developed for MK9 - 5 cigarette machine , and thetests werecarried outin thelaboratory .It has been shown thatthe new intelligent diagnosticexpertsys tem based on ANN has higherliability than the com mon diagnostic expert system .