Application Research Of Exhausting Gas Measure And Fault Diagnosis Of Electrical Control Engine

With rapid development of automobile industry, the exhausting gas from the vehicle brings out severe influence on the living environment.Especically, the exhausting gas of the used vehicle that do not accords with the regulated standard pollutes seriously our living environment.The article discusses the purpose and significance of the subject. it is dictated definitely that if we depend on the exhausting gas from engine to diagnose the fault of automobile by applying the neural network model.the result reduces the contamination of the exhausting gas of the used vehicle on our environment, abridges repair time, decreases the repair master's labor intensity and improves labor efficiency.Meanwhile,it is in favor of the management of the I/M system.The article mainly research that the BP(Back Propagation) and the RBF(radial Basis Function) neural network is applied for the fault diagnosis of the electronic control engine.First,This paper introduces the contamination of the exhausting gas of the vehicle, the I/M system and the basic knowledge of the electronic control engine.Second,This paper introduces the configuration and arithmetic of the BP and the RBF neural network,and buildes the neural network model.Last,taking Lexus400 engine as example, simulating almost all kinds of engine faults in the idle condition. These exhausting gas is gathered by using the JinDe K81 and the NHA-501 equipment.moreover,building the stylebook.Using MATLAB language and data management technology to compile simulation programme. this thesis traines and simulates to the stylebook by adopting the BP neural network and the RBF neural network and compares the result of the training and simulation.The conclusion that the training rate of the RBF neural network is faster than the BP neural network,the efficiency of the RBF neural network is higher than the BP neural network and the diagnosis result of the RBF neural network is more correct.Thus, the RBF neural network for the malfunction diagnosis of the electronic control engine is more feasible and successful than BP neural network.