Research Of Radar Object Recognition And Tracking Technology Based On Trolleybus Collision Avoidance

In recent years,many large and medium-sized cities vigorously promote the public transport system.With the concept "green traffic" putting forward,trolleybus,as a new energy public transport,gradually revives.However,the width body and big turning radius,leading to a lot of visual blind area and traffic accidents.Laser radar can detect the car situation,identify the size of the obstacles and motion state of obstacles.Applying the laser radar technology to the trolleybus are helpful to reduce the labor intensity of drivers,so that it can improve vehicle safety,reduce road traffic accidents and improve the efficiency of traffic system.(1)This paper analyzes the development situation of trolleybus,elaborates the obstacle target recognition and target tracking technique respectively.It analyzes the function demand and system overall design of trolleybus obstacles target recognition and target tracking and especially explains the target partition clustering and target tracking.The partition clustering uses the Grid algorithm to cluster the laser radar scatter and set up the front target classification.Target tracking uses Extended Kalman Filtering algorithm and nearest neighbor algorithm combination to discriminate the clustering results and track the target obstacle.This paper mainly studies laser radar target recognition and tracking technology route that can be applied to the trolleybus collision by the relevant algorithm.(2)This paper elaborates the structure of the trolleybus obstacle recognition and target tracking system,then explains the information acquisition,target recognition and target tracking module respectively.It also designs the function and structure of laser radar data collection software and laser radar real-time display software,then completes the software by C++ programming language.(3)The paper conducts real vehicle testing and verification of the trolleybus obstacle recognition and target tracking system,then the experimental results are analyzed.By comparing the real car gathering information with test plan information,it verifies the effectiveness of Grid algorithm,extended Kalman filter algorithm and the combination of nearest neighbor algorithm.By testing the function of the software,the paper realizes the car environment information collection,real-time radar display,real-time video display and the front obstacle clustering processing.At last,it proves the practicability of the software.The experimental results show the trolleybus obstacle avoidance and target tracking system proposed by the paper can accurately capture the environmental information of the vehicle.It can realize obstacle identification and target tracking,and provide real-time and accurate vehicle warning information for the driver.