齿轮故障诊断专家系统

Expert System for Diagnosing Gear Faults

在分析了以往的齿轮故障诊断方法及大量试验数据的基础上，选出了对应于各故障敏感的参量，总结归纳了齿轮故障的判据并在此基础上建立了故障诊断的专家系统，把模糊理论用以知识表示进行不精确推理；采用元知识控制下的反向推理机制并根据推理结果赋以合理的解释，该系统既能本机自动诊断又能结合其它诊断方法作精密诊断；还能兼顾在线监测与离线分析。利用该系统对齿轮试验数据进行分析、处理进而推理、诊断获得了满意的效果。

Under theoritical study and analysis signal of gear vibration by various processing methods, some effective criteria are summed up and the expert system is set up. In this system, in order to make uncertain reasoning, fuzzy theory is applied to knowledge presentation;the backward reasoning mechanism controlled by meta-knowledge is adopted; the reasenable explanation according to the reasoning result is attached as well, The other features of this system are: 1) not only being useful for automitic diagnosis by itself but used for pricise diagnosis cooperated with other methods, and 2) having consideration of both on-line monitoring and off-line analysis at the same time.