摘要
本文提出一种兼具推理及运算功能的故障诊断专家系统的构想.图论是分析设备故障与症状对应关系的有效方法.以此为基础,在PROLOG的支持下进行命题演算及谓词演算,构成推理机,与基于模糊集理论的运算器进行信息交换,组成诊断专家系统.以柴油机为对象阐述了专家系统诊断故障的推理过程,并举出应用最大生成树和Fuzzy ISODATA两种模糊聚类方法诊断柴油机故障实例.将FuzzyISODATA方法与统计模式识别方法的诊断结果进行了比较,两者是一致的.
The construction of an expert system, which can do the work of calculation and inference, for fault diagnosis of the diesel engine is discussed. Fuzzy graph is a useful tool for analyzing relations between troubles and symptoms of equipment. By using predicate calculus and with the help of PROLOG exchanging information with ADSDE,the inference engine of the expert system can successfully achieve our purpose. Clustering analysis methods by means of fuzzy graph and fuzzy ISODATA are introduced. Examples of trouble shooting of the diesel engine by fuzzy ISODATA, compared with the result of stastistical pattern recognition method,are described.
关键词
柴油机
故障诊断
专家系统
模糊集
Expert systems
Fault diagnosis
Diesel engine
Graphtheory
Fuzzy logic
Fuzzy clustering analysis
