Traffic Flow Monitoring And Identification Based On Millimeter Wave Radar

With the continuous development of our economy,the travel modes of citizens are becoming more and more diversified.At the same time,the traffic pressure has increased sharply,with the continuous flow of different vehicles in the city.To improve the traffic environment as much as possible,the millimeter-wave radar based on a linear frequency modulated continuous wave traffic monitoring system is proposed to monitor all-weather and real-time traffic intersections or toll gates and other high-accident sites.Intelligent monitoring is selected to replace the previous manual monitoring.Monitor the vehicle’s speed,track the vehicle’s position,measure the size of the vehicle to determine the vehicle type,etc.through intelligent traffic,and optimize the traffic environment to improve security.This paper briefly describes the development advantages and characteristics of millimeterwave radar and the current research status of the intelligent traffic monitoring system.Compared with the video monitoring system,the millimeter-wave radar monitoring system can work normally in all-weather,and measure the distance and velocity correctly.In this research,we analyze the distance and velocity decoupling method of linear frequency modulated continuous wave signal,realize the improvement of angle resolution through the expansion of MIMO radar virtual array,study the angle estimation method of array signal,and improve the three-dimensional target clustering algorithm to achieve the purpose of target clustering and tracking.In the original signal which is obtained by the chirp continuous wave signal,the reason for the coupling between the target distance and the velocity is analyzed,and the two-dimensional spectrogram of the target distance and speed is obtained by using the two-dimensional FFT method,which is used for solving the distance and velocity of the target from the radar.The principle of the array element virtualization technology of MIMO radar is analyzed,and the time-division multiplexing method of MIMO radar is selected to expand the virtual array.The angle resolution capability can be effectively improved by using the MIMO virtual array through simulation analysis.For the estimation of the target angle information,the theory method and simulation analysis are carried out on several angle estimation methods of the array signal.Based on the problems of conventional two-dimensional clustering algorithm in detecting vehicle targets,and the analysis of vehicle point cloud distribution,a three-dimensional clustering method with the ellipse as the neighbor is proposed,which improves the accuracy of the multi-target classification and realizes the target point cloud tracking through the tracking algorithm.The millimeter-wave radar system in this paper is based on the TI series millimeter-wave radar sensor module IWR6843 ISK.A set of signal processing procedures are designed and the data acquisition board with configuration parameters is used to collect raw data.The processing of single-target data and multi-target data are analyzed respectively,and the tracking results of the target are achieved and the size of the target vehicle is identified.