计算智能在故障诊断专家系统中的应用

基于知识的专家系统及其在故障诊断中的应用研究十分活跃，形成一类故障诊断专家系统，但该方法存在知识获取的瓶颈、推理能力弱及难于处理不确定知识等问题，其进一步发展受到严重的阻碍。与此同时，神经网络、模糊逻辑和遗传算法等计算智能在近年来却取得了一系列的进展，并具有专家系统等符号智能所缺乏的优势。将计算智能应用到专家系统中即与符号智能相结合起来能有效地解决诊断专家系统的难题，该文对此进行了探讨，并提出了一些新的思路。

The Application of Computstional Intelligence in Fault Diagnostic Expert Systems

The study of expert system based on knowledge and its applications in fault diagnosis is very active,and forms one kind of fault diagnostic expert system. However, this approach exists some problems to hinder itsdevelopment such as the bottleneck Of knowledge acquisition, weakness of reasoning ability and difficulty inmanaging uncertain knowledge etc. At the same time, the computational intelligence which includes neural networks, fuzzy logic and genetic algorithms, obtains a series of developments and has some advantages, which thesymbol intelligence (such as expert system etc.) lacks. The combination of computational intelligence and symbolintelligence can effectively solves the poblem in diagnostic expert system. This paper discusses the issue andpresents some new ideas.