轴承事故的神经网络（ANN）分析

用神经网络分析和模拟轴承事故，分析了基于Ｐｅｔｒｉ网的事故分析的优点和不足。通过构造事故的神经网络，可求出导致顶上事件发生的最小割集和控制事故发生的最小径集，这种分析方法为神经网络在故障诊断中的应用开拓了新的研究思路。

THE BEARING ACCIDENT ANALYSIS BASED ON ARTIFICIAL NEURAL NETWORKS

in this paper,The bearing accident is analysed by using artificial neural network. The merits and defects of accident analysis based on Petri net are analysed. The minimal outset leading to the occurrence of the topaccident and the minimal pathset controlling to the occurrence of the accident can be found by structring neural network of the accident. The method opens up a new branch for the application of artificial neural networks in fault diagnosis region.