基于模糊专家系统的评估诊断方法

Fault Assessment/Diagnosis Based on Fuzzy Expert System

针对设备故障诊断中存在的不确定性问题,提出一种将模糊逻辑和专家系统结合的评估诊断方法·模糊专家系统中采用模糊产生式规则表示知识,通过采用模糊匹配的不确定性推理方法进行,并采用以置信度高低为标准的冲突消解策略来实现基于知识的诊断方法·文中介绍了专家系统功能组成和实现方法,同时在电力变压器的气相色谱分析故障诊断方法中应用了该模糊专家系统·通过现场近百次实际诊断,准确率可以达到95 7%,验证了将模糊专家系统应用到变压器的故障诊断中是可行的·

A method of fault assessment/diagnosis that combines fuzzy logic with expert system is put forward to the uncertainty problems to be solved in equipment fault diagnosis. It is done the way the rules originated from fuzziness are introduced in fuzzy expert system to express knowledge obtained with fuzzy matching procedure adopted in the uncertainty reasoning process to implement the diagnosis based on knowledge through a strategy to resolve conflicts with the confidence level taken as criterion. The functional components and implementation of the fuzzy expert system are also introduced,which has been applied to the fault diagnosis using gas chromatography for power transformers. The percentage of accuracy of the method is up to 95.7% as proved by the practical diagonosis done in site near 100 times, thus verifying the fact that applying the fuzzy expert system to the diagonasis for power transformers is applicable.