| With the extensive application of vehicle automatic transmission, fault diagnosis as a key technology of automatic transmission, can effectively improve the maintenance efficiency and vehicle driving safety. AT is a complex system which including mechanical subsystem, electronic subsystem and hydraulic subsystem. Its configuration and theory is complicated, which results in difficultly to diagnose.The research on fault diagnosis method of automatic transmission is the foundation for the development of more powerful diagnostic system. The principle is establishing the model of the automatic transmission, and making full use of the existing sensor information as much as possible. The mechanical and hydraulic faults could be diagnosed by a real-time fault detection and diagnosis method of the AT. With the development of diagnostic technology and further study in AT failure analysis, the real-time monitor system for fault diagnosis by using computer technology has the important practical significance to expand the scope of the existing fault diagnosis system.Based on failure analysis of mechanical transmission system and hydraulic control system of the automatic transmission and the premise that the fault can be diagnosed, a real-time fault detection and diagnosis method of the AT is proposed.The logic analysis method of the automatic transmission is put forward,for the steady state fault, based on the relationship between the input shaft and output shaft speed signal is presented.For the transient fault of the automatic transmission, the fault diagnosis method of observer is presented based on the instantaneous redundancy relationship between the input shaft and output shaft speed signal.Modeling of AT in AMESim. Study on qualitative simulation for some performance parameters of AT is carried out. Steady state and transient AMESim simulation model are established.Two cases on qualitative simulation when the typical part works normally and part is disabled are researched by using batch run function of AMESim. Failure analysis is validated by compared with results of simulation. Much sample data for fault diagnosis are provided.Then the fault detection and diagnosis method based on support vector machine is proposed for automatic transmission. After data processing, fault diagnosis model is acquired, and the fault diagnosis model is established to support vector machine to detect and diagnose the fault of automatic transmission.At last, the fault detection and diagnosis algorithm based on the model of fault detection and fault detection for the automatic transmission is proposed. A pressure observer is designed, and the simulation model of AMESim is built to verify the correctness of the design of the clutch pressure observer. With the oil pressure estimation observer, the fault of the automatic transmission shift solenoid valve is an example to realize the transient monitoring and fault diagnosis of the automatic transmission system. |