Research On The Key Technologies Of Preceding Vehicle Collision Warning System

With the increasing number of car ownership in China and the continuous expansion of highway construction scale,vehicle rear-end collision on expressway has become more frequent,timely and effective vehicle anti-collision warning is an important means to prevent the occurrence of it.In this paper,the preceding vehicle collision warning system is taken as the research object.The machine vision image processing technology and ZigBee transmission technology are applied to study the key technologies such as vehicle detection ahead,vehicle distance measurement,vehicle-vehicle communication system and automobile anti-collision warning model.Firstly,in order to reduce the influence of excessive interference factors in the forward vehicle detection process and improve the real-time and accuracy of vehicle detection,this paper is based on highway road design specifications and driving environment characteristics,the ROI of the region of interest is set for the vehicle image acquired by the vehicle camera.The ROI is detected by the forward vehicle detection algorithm based on Adaboost and Haar features,and the vehicle detection result is obtained through prior knowledge.Vehicle presence verification is performed to provide accurate vehicle images for later vehicle ranging.Secondly,according to the characteristics of the car license plate with blue background and white typeface in China,the license plate location algorithm based on color feature and projection information is designed.By locating the license plate,the width information of the license plate image is obtained.And the distance measurement to the preceding vehicle is completed according in the license plate size fixed and the camera imaging principle.Then,the vehicle-vehicle communication system based on ZigBee technology is built.The front car speed,brake pedal state and the ego vehicle speed are collected in real time by utilizing the VK162 speed sensor and the SG-230 pedal stroke measuring instrument.The Xbee-Pro S3 B communication node and the Arduino Uno R3 development board are used to establish a vehicle-to-vehicle communication module.Finally,on the basis of traditional TTC warning model,the emergency braking condition of the preceding vehicle is simulated in the PreScan8.5.0 and MATLAB/Simulink platform,and the safety anti-collision warning time threshold under the emergency braking condition of the preceding vehicle is obtained.The least squares method is used to fit the simulation test results to construct a vehicle anti-collision warning model based on vehicle-vehicle communication and vehicle driving state.In order to verify the reliability and accuracy of the model,a preceding vehicle collision warning system based on vehiclevehicle communication and visual perception technology is built in PreScan.The early warning model proposed in this paper is compared with the traditional TTC warning model under the emergency braking condition of the front vehicle and the test results were analyzed.Through the research on the key technologies of the preceding vehicle collision warning system,it is verified that the comprehensive use of machine vision image processing technology and vehicle-vehicle communication technology can improve the accuracy of the preceding vehicle collision warning system under emergency conditions.It is of great practical significance to improve the driving safety of vehicles and avoid the occurrence of highway rearend accident.