Application of fuzzy Petri net to knowledge representation and inference in fault diagnosis

The conventional fuzzy production rule is complicated and lowly efficient in knowledge representation and inference,and the single threshold value can not meet the requirement of fault diagnosis.Therefore,a general method and its corresponding fault diagnosis model were established for transforming fuzzy production rule to fuzzy Petri net.And meanwhile the two-level algorithm was presented based on setting double threshold value.The typical method can be used not only to analyze and diagnose the three kinds of states of the system and complex faults effectively,but also to handle faults in the remote control system of a certain diesel.The results indicate that the speed and efficiency of inference can be improved by setting the double threshold value,thereby detecting and removing potential faults.