| The tank is a type of weaponry with special use and very atrocious working condition, higher safety and reliability on demand. Moreover, the fuel system of diesel engine isn't only the hard core of affecting on the vehicle maneuverability, but also is the main cradle land of mechanical fault. So it is important practical significance to carry condition monitoring and prediction for the tank fuel supply system into execution.The paper attempts to research profoundly fault mechanism, fault feature extraction and fault diagnosis method of diesel engine. Its research object is the fuel system of tank diesel engine and its academic foundation is test technology, signal processing, artificial neural network, etc. A fault diagnosis system of the diesel engine based on it is built up. During the process the following tasks are accomplished.(1) Combining the fault mechanism and the pressure waves of the tank fuel supply system, we researched on the method of pressure signal processing and analysis in magnitude domain and frequency domain. A extract method of fault feature extraction was put forward.(2) Using export system for reference, we bring forward the principle model of condition monitoring parameter optimum select. According to the frequency of different abnormal condition and its degree of danger, weighting coefficient was introduced, feature parameter whole sensitivity and stability evaluation index was established.(3) Base on the rapid algorithm of artificial neural network(ANN), and ANN is used to realize automatic fault classification and fault diagnosis for the fuel supply system of diesel engines, a fault diagnosis system is designed by using LabVIEW software. The test is carried out on a running diesel engine, and the satisfactory results are obtained.(4) Using Virtual Instrumentation for reference, a fault diagnosis system is designed based on Virtual Instrumentation, the repeatability of results proved practicality of this system.
|
PDF Full Text Request