Research On Typical Fault Analysis And Diagnosis Technology Of Marine Diesel Engine

As one of the main motive power of marine integrated electric propulsion system, the performance of diesel engine has great influence on the performance of the whole electric propulsion system. Due to the complex structure and poor working environment of diesel engine, so there is a great possibility of failure. Once a fault occurs, not only will the diesel engine's performance be greatly reduced, but also will make the propulsion system performance decrease, and even cause accident. Therefore, the monitoring and fault diagnosis of diesel engine is of great significance. This paper takes the diesel engine for the power generation in the ship integrated electric propulsion system as the research object. The dissertation develops from the following aspects: the failure mode of diesel engine, diesel engine fault features, fault diagnosis, condition monitoring and fault diagnosis software development. The main research contents are as follows:(1) Analysis of diesel engine failure mode. Firstly, the working principle and the composition of the diesel engine are briefly introduced, and the common faults of the cooling system, the gas distribution system and the fuel system are summarized. And the failure mode and effect analysis method is used to analyze the faults of the above three systems, then the mechanism and influence of various faults of diesel engine are obtained. Finally, the failure mode and analysis table of diesel engine's cooling system, gas distribution system and fuel system is obtained. Based on that, five failure modes are identified to study in this paper.(2) Fault characteristic analysis of diesel engine. First of all, the mathematical model of 16 cylinders diesel engine is established. On this basis, the mean value model of the diesel is built. In view of the common five kinds of faults of diesel engine: the air filter clogged, cooler gas side fouling resistance and the cold water side fouling, turbocharger failure and turbine protection grille jam fault, are simulate in model respectively, under the load of 100%, 90%,80%, 70%, 60%, 50% and under different degree of fault. Through the simulation of fault on the simulation model, the changes of diesel engine performance at each fault are obtained.(3) Diesel engine fault diagnosis. PNN neural network and PNN neural network based on PCA are respectively used to establish 16 cylinder diesel engine fault diagnosis model. The results showed that PNN neural network based on PCA has higher efficiency and 100%diagnostic accuracy. At the same time, taking exhaust temperature as the research object,using the grey theory to predict the fault of the diesel engine,which provides the technical support for the health management of the diesel engine.(4) Condition monitoring and typical faults diagnosis software for diesel engine. On the basis of establishing the fault diagnosis network of diesel engine,the condition monitoring and typical faults diagnosis software of diesel engine is built. The condition monitoring and fault diagnosis of the 16 cylinders diesel engine under 70% load are carried out. Finally,the monitoring and typical fault diagnosis of diesel engine is realized. At the same time, the software is used to realize the failure prediction by exhaust temperature. The results show that the software can predict the fault.