The Application Research Of Optimization Algorithm For Intelligent Fault Diagnosis In Ship's Diesel Engine

The thesis does detailed research in the optimized algorithm applying in the ships diesel engine intelligence breakdown diagnosis. The intelligent breakdown diagnosis system is the fuzzy neural network which is optimized.The ships diesel engine is a complex system, which is the power fountainhead of the entire ships. Its normal operation has the important significance to the ships work security. Because it involves many domains of the machinery, the thermal energy, the signal examination, the safekeeping of security, the control and so on, and its work status is close correlation with a series of subsystem just like fuel oil, turbo-charged, burning, cooling, lubrication and so on which is the system of the multi-breakdown attribute. Therefore the question of the breakdown diagnosis needs to do the synthetic study from different angle. Firstly the breakdown theory and the breakdown diagnosis process of the ships diesel engine are introduced in the thesis and the characteristic parameter of the ships diesel engine breakdown diagnosis is given in the research of the thesis.In recent years advanced breakdown diagnosis technology develops swift and violently specially in many methods and technologies of intelligent breakdown diagnosis, which has provided the powerful guarantee for solving the breakdown diagnosis of complex system. In the intelligent breakdown diagnosis, in the progress of the breakdown indication vector to the breakdown vector continuously mapping, systematic and forms fuzzy neural network (FNN) by unifying the artificial neural networks and the fuzzy logic, which has the merit of fuzzy logic system effectively using fuzzy information, and also has the characteristic of parallel processing of the neural network and highly studying information. The fuzzy neural network overcomes the insufficiency of the sole system, which can be better adaptation in the intelligent breakdown diagnosis system. The intelligence breakdown diagnosis system of the neural network will be application in the ships diesel engine and do research of simulation in the thesis. Finally, this thesis, for the first time, aim at the diesel engine breakdown diagnosis system and combine with the genetic algorithm and the ant group algorithm to train the parameters of the fuzzy neural network. After simulating the intelligent breakdown diagnosis system of the ships diesel engine and comparing with the BP algorithm, we get better diagnosis result and enhance the ability of the system. Also, it converges fastly, and overcomes the shortcomings of the BP algorithm and get into extreme value of network training. All of the conclusions demonstrate the feasibility and validity of the researching work.