Study On Vehicle Forward Collision Avoidance System

Today traffic accidents have become a social problem all over the world and many researchers are paying remarkable attention to the advanced safety technologies for improving the situation. As one of the important technologies, the vehicle collision avoidance system based on ITS(Intelligent Transportation Systems) has been investigated in many countries. However, till now some key problems in this field haven't been well resolved. So that the more theoretical and technical study on vehicle collision avoidance system will be essential to improving road traffic safety, decreasing traffic accident rate and promoting ITS application in China. In this dissertation, study on vehicle forward collision avoidance system has been conducted based on vehicle system dynamics and control theory. The study covers with the system dynamics modeling and control algorithm, the estimation of safety distance, the acquisition and process of vehicle and environmental information and the development of the vehicle test system.Firstly, the outline of the vehicle forward collision avoidance system is designed in this dissertation. Based on it, the key technologies involved in the system and their relationship are determined.As the basis to realize and evaluate the system function, the dynamics modeling of the collision avoidance system is studied. According to the dynamic characteristics of the test vehicle system and the actuator structure, a longitudinal dynamics model which can be used to simulate the dynamic characteristics of the collision avoidance system under some complicated working conditions has been developed. Furthermore the veracity and simplicity have been compromised in the model.In order to obtain the safety distance of moving vehicle in real-time, a new safety distance model is proposed. The model is established on the intention assumption of drivers' distance keeping maneuver in actual traffic situation and the model parameters can be identified by statistical process of many skilled drivers' operation test results. Simulation and test results show that the proposed model represents the driving characteristics of human driver well and can be applied to different kinds of traffic situations.Based on the research mentioned above, the control system of the collision avoidance system is developed. The control function defining layer that can change the function automatically and the human simulated robust controller of vehicle collision avoidance system are implemented. The control function defining layer is designed on the foundation of the system function definition of the vehicle forward collision avoidance system and the driver's manipulation prior principle. So the layer can make the decision of reasonable control function under different situations. The robust controller of vehicle collision avoidance system consists of a lower-layer controller, which is designed based on MMC (Model Matching Controller) method with a H∞ compensator, and an upper-layer controller, which is designed based on LQ (Linear-Quadric) optimal control theory and the manipulation characteristics of human drivers. Utilizing the developed control system, the vehicle dynamics system can be controlled to realize the function of forward collision avoidance in the way keeping accordance with the human drivers' manipulation characteristics under many complicated traffic situations.To acquire the environmental information under different road traffic situations, the target vehicle identification and its moving state information obtaining method are investigated. The target vehicle identification method is developed on the basis of the domestic road construction standard, which can effectively identify the target vehicle from complicated traffic situations. The moving state information obtaining method of target vehicle is proposed with the four-state Kalman filter, and it can quickly and exactly obtain the moving state information of the target vehicle. Moreover, on the basis of the statistics theory, the target vehicle tracking meth...