Fuzzy Petri-Nets Based Fault Diagnosis for Mechanical - electric Equipment

With lots of incomplete and uncertain information, a highly effective inference model is needed to build a fault diagnosis expert system to monitor the manufacturing equipment. Based on the inverse driven information flow, a new fuzzy Petri-net model called FFDPN (fuzzy fault diagnosis Petri-nets) is presented in this paper for the fault diagnosis of mechanical -electric equipment. The production rules are defined backward, and the diagnosis model is more rigorous and more effective than the general fuzzy Petri-nets. The firing rules of transitions in FFDPN are defined. The method of knowledge representation and the inference algorithm are also proposed. The model can be used to model a fault diagnosis expert system. A fault diagnosis example using FFDPN for a diesel engine is given to test the method. It is demonstrated that the diagnosis inference of this model is quite effective.