Bp Neural Network-based Driver Fatigue Research

The hazard of driver fatigue is more and more serious on the traffic safety. The research has indicated that about twenty-five percent to thirty percent of the traffic accidents were caused by the driver fatigue. The driver fatigue influences the driver's vigilance and the ability for safety-driving. The problem of the drive fatigue is becoming much cause for concern. The Western developed countries have devoted large amounts of human, financial and material resources to carry out research on the driver fatigue. For the time being, the research of the driver fatigue is relatively backward in our country. So the significance to the research of the driver fatigue is obvious.The paper introduces the theory of fatigue and the monitoring method of driving fatigue.Based on them,first, from the perspective of automatic control ,it gives a quantitative analysis to the formation of the driver fatigue, and establishes the mathematical model of the driver fatigue;Then, followed by subjective survey and the corresponding test methods, it analyzeds impact between the driver fatigue and drivers'self factor such as age,sex and so on which have little effect on driving fatigue.The paper, using the AHP improved method,gives a hierarchical analysis on the driver fatigue symptoms, and confirms the weight of various of fatigue symptoms.According to the weights, it extractes the rotation angle and the reaction time which are the input of BP neural network model.Through analysis, it confirms the parameters of driving fatigue BP neural network model. While,using the relevant test data,it establishes the evaluation model of the driver fatigue which is the comparison of the model of the driver fatigue BP neural network. Finally,using GUI in Matlab, it initializes, trains and simulates the model of driver fatigue BP neural network.It indicated that the state of the driver fatigue was estimated by the Error Back-Propagation Algorithm Network model to the driver fatigue better, through compared the results of simulation with the results of forecasting.