Highway Vehicle Rear-end Collision Probability Model And Simulation

Rear-end collision of express high-way is a serious traffic accident and has caused extensive attention of the transport sector in recent years, people has been trying to solve the problem of rear-end collision from the prospect of technology and theory. And the mathematical model of rear-end collision that established in the process is considered to be the basis of studying this phenomenon.In order to improve the Active safety performance of vehicles under the circumstances of high speed driving, first, it analyzes and discusses the accident mechanism and predisposing factor and its relevance, and then it analyzes the Kinematic characteristics of the following car and the front car in the driving condition. According to the calculation method of braking distance of automobile theory, different calculative models of safe distance are deducted in connection with different movement condition. And also studies the key parameters's stochastic and dynamic nature of the model, proposed a fuzzy-based determined method for the braking lag time. On the basis of the above analysis, the concept of probability of rear-end for vehicles. The safety condition of vehicles are affected by a lot of factors, the seriousness of the effect can be reflected by macro risk posibility, that is the possibility of rear-end collision, the possibility of rear-end collision in different areas are concluded according to the safety distance models and the change of speed as well as the weather conditions to establish the various factors that influencing the rear-end collision.Taking advantage of VB Software programming to simulate the state of the vehicles, drawing numbers on some constraints factors randomly, carried out a lot of experiments and finally analyze the data for model in comply with the actual effect of the high way. Based on the based on neural network theory, and establish the BP network possibility model for rear end collision, and consider such factors as horizon, after the vehicle speed, the speed difference between vehicles, road adhesion coefficient, the weather forecast to be the influencing factors and safety, rear-end collision as a model of the probability of outcome indicators, ultimately the neural network model we received reveals the relationship of the rear-end collision of high way in a certain error range, which can be applied to the rear-end forecast for real-time condition.