| As the environmental pollution and the shortage of resources are getting more severeover the years, auto industry turns to the development of new energy vehicle. In recent years,hybrid electric bus has been widely used under the support of the government. Meanwhile,Automated Mechanical Transmissions (AMT) is largely adopted on hybrid electric bus dueto its high efficiency, good inheritance of production line and low cost.Compared with conventional transmissions, due to the presence of electrical controlsystem, the structure of AMT becomes more complicated and the failure rate is increased. Asa result, the fault diagnosis of AMT control system has great significance to the vehicledriving safety and vehicle maintenance. The AMT fault diagnosis system for hybrid electricbus is researched in this paper.First, the goal and main content of the research is determined by summarizing theresearch and development status of hybrid electric bus and fault diagnosis of AMT. Then, thestructure, working principle and working mode of the AMT system for hybrid electric bus isdescribed and its fault analysis is conducted. Through the analysis, a hierarchical faultdiagnosis strategy is adopted because it is difficult to use a single fault diagnosis method torealize the fault diagnosis of the AMT system. After that, the detail fault diagnosis strategyof every layer is introduced, including the preliminary fault diagnosis strategy of sensors, theanalytical redundancy method, factor tracking method and the deep fault diagnosis strategyof sensors based on wavelet neural network algorithm. Among these fault diagnosisstrategies, the focus is the analysis of the deep fault diagnosis strategy of sensors based onwavelet neural network algorithm.Finally, the adopted fault diagnosis strategy is tested. The tests mainly include thesoftware simulation and hardware simulation: software simulation is carried out inMatlab/Simulink simulation platform. The vehicle model and fault model are built, and thenthe AMT fault diagnosis module is established and tested through typical faults to verify the effectiveness of the fault diagnosis strategy of every layer. In hardware simulation, firstly,the hardware simulation platform based on MotoTron is built. Then the established controllermodel and vehicle model are downloaded to MotoTron and target machine respectively.After that, the function of the fault diagnosis module is verified using the trouble-free modeland failure model. The effectiveness of the proposed fault diagnosis strategy is verified by aseries of software and hardware simulation tests. |