A simulation-based expert system for process diagnosis

Expert systems have been widely applied to a variety of domains including the field of intelligent industrial diagnosis. However, knowledge acquisition has been often considered as the bottleneck in expert system development. This paper describes the architecture of an expert system in industrial process diagnosis, including simulation-based knowledge generation, fuzzy knowledge representation and case-based diagnostic reasoning. In this paper, a methodology of acquiring knowledge and building up knowledge bases based on software simulation is introduced. The efficiency and accuracy of the simulation-based methodology had been verified by experimental results of performing such an expert system in a hydraulic brake system.