某型监控系统故障诊断专家系统知识获取和故障树推理分析

Knowledge Acquisition and Reasoning Analysis of A Certain Type of Monitoring System Fault Diagnosis Expert System

基于故障树技术的专家系统适用于如某型船柴油机监控系统的复杂容错系统故障诊断。知识的获取和推理中的故障树分析是专家系统关键部分。首先将监控系统各故障事件间逻辑关系用故障树的形式联系起来,然后转化成二叉树的形式存储故障知识。对专家系统推理机中复杂故障树用线性搜索方法模块化并查找其静态子树和动态子树。静态子树用二元决策图进行分析,动态子树用马尔可夫模型进行分析。通过故障树分类分析,得到准确全面的推理结果。

Fault diagnosis expert system（ES） based on fault tree technology is suitable for complex fault -tolerant system such as a certain type of ship monitoring system. Knowledge acquisition and fault tree analysis in inference engine are key parts in ES. Firstly, the logical relationship between types of ship monitoring system fault events with fault tree form in series are connected Then in inference engine of the ES？ a binary tree is used to store the fault knowledge, linear search method is used to find the static and dynamic sub -trees in a complex fault tree. Static sub-tree with binary decision diagrams and dynamic sub -tree are analyzed with Markov Model, accurate and comprehensive reasoning result are gotten by classified fault tree analysis.