故障诊断专家系统神经网络学习机研究

A STUDY OF THE ANN LEARNING MACHINE OF FAULT DETECTING EXPERT SYSTEMS

根据故障诊断专家系统的特点，设计了一种面向故障诊断问题的机器学习系统。在参数修正自学习模型中，引入了变结构的参数修正人工神经网络模型，采用ＢＰ学习算法及后验学习方式，实现了故障诊断专家系统参数修正学习机设计。在汽车故障诊断问题中的应用结果表明，学习效果良好。

The paper presents a study of the artificial neural networks (ANN) learning machine which is based on the fault detecting expert system. A variable architecture ANN model with B P algorithm is introduced to the learning machine of the expert system. Some parameters can be automatically modified in the system.