基于模糊概率Petri网的故障诊断建模与分析

Modeling and analysis of fault diagnosis based on fuzzy probability petri net

为筛选故障诊断推理中的冗余信息,研究了面向故障诊断的模糊概率Petri网(FPPN)。通过综合专家经验知识,将事件的概率参数引入推理分析中,并给出了基于状态方程的正向、反向及混合推理分析方法。以铁路道岔系统的故障诊断为例,说明基于FPPN的建模和分析方法,结果表明:引入概率参数的FPPN能够筛除冗余信息,分析过程体现了人工推理的思维模式,优于基于FPN的分析结果。

In order to filtrate some unnecessary information in fault diagnosis reasoning process, we have studied the fuzzy probability Petri net （FPPN） for fault diagnosis. Via colligating experiment knowledge of experts the paper used the probability of every event in the reasoning process and presented the forward, reverse and mixed reasoning methods based on state equation. An example of railway turnout system fault diagnosis indicates that the modeling and analysis methods based on FPPN are feasible. The result shows that they can filter unnecessary information better than those based on FPN and that the reasoning process embodies man＇s thinking.