Research on fault diagnosis of electric appliance for vehicle based on CAN bus

Aimed at the problem and difficulty of diagnosis after the event in the traditional form of the fault diagnosis of electric appliance for vehicle, a new fault diagnosis system is presented which is based on the combining of CAN bus and embedded database. Using a probabilistic neural networks (PNN) algorithm, the state parameters of engine were obtained by CAN bus and were used to identify multiclass-state with the samples prestored in the embedded knowledge base, and with the exceptional signals from the sensors, a contingent fault can be estimated and determined effectively. The algorithm has shorter training time and higher right diagnostic level compared to back-propagation neural networks. A method of the fault diagnosis of self-determination and intellectual for vehicle is realized, as the result, the handling performance of vehicle is improved, and the probability of fault present can be decreased, so does the maintenance cost.