On the Intelligent Fault Diagnosis Method for Marine Diesel Engine

The marine diesel engine is a complex system. Its mapping process of fault diagnosis has multi-fault attributes, which means input and output of fault pattern attribute are the multi-mapping relations. An approach of intelligent fault diagnosis using fuzzy neural networks and genetic algorithms to optimize and train is studied in this paper for this system. The structure and the model of intelligent fault diagnosis made up of fuzzy neural network were introduced. Its weight and the threshold value optimized and trained by the genetic algorithm are presented. Finally, this fuzzy neural network system optimized and trained by genetic algorithm was applied to the fault diagnosis of the marine diesel engine. The simulation showed feasibility and validity of this method. The precision of fault diagnosis can be improved effectively, and the generation capacity of the intelligent fault diagnosis system and the accurate knowledge expression are enhanced.