Design And Simulated Implementation Of Vehicles' Frontal Collision Warning Algorithm

The issue of traffic safety is a perpetual theme. With the rapid increase of the number of the vechicles, the number of traffic accidents each year is always high and vehicles'safety problem has become an increasing focus. Vehicles'frontal collision warning system could avoid collision with frontal vehicle by offering sound or light warning signal, but the former vechicle's frontal collsion warning algorithm (VFCWA) haven't achieve this aim. Furture more, although drivers have different requirements of VFCWA, study on this point is far from engouth. So designing a reasonable vehicles' frontal collision warning algorithm has great value to improve the vehicle's initiative security and to meet the needs of driving-behavior.Safety distance model is the basis of the frontal collision warning algorithm. After analyzing and amending the former models, an effective safety distance model has been proposed in this thesis. In addition, because the vehicles'frontal collision warning algorithm should also adapt to the driving-behavior characteristic, a driver's braking-behavior model (DBBM) using BP nerural network has been designed to predict the driving-behavior. Then, the new frontal collision warning algorithm is founded which combines the safety distance model and the braking-behavior model.Firstly, several existing schemes on frontal collision warning algorithm are introduced. After analyzing the process of vechicle's breaking, an improved safety distance model is proposed, which solves the disadvantage that is not enough accaute of the former safety distance models. Secondly, based on the adaptability requirement of collision warning algorithm and the knowledge of driving-behavior theory, a method which uses driving-braking behavior to describe the driving-behavior characteristic in dangerous situation is proprosed and designed.Then, the DBBM is founded by BP neural network. Afterwards, VFCWA reflecting driving-behavior is built, which combines the safety distance model and DBBM. In order to verify the built collision warning algorithm, a virtual driving system in VC++ using OpenGL API is implemented, and the simulated implementation is accomplished in this virtual system. Finally, the built collision warning algorithm is analysed and summarized, and some disadvantages of the algorithm are pointed out for further improvement.