Research On Vehicle Classification Recognition And Tracking Technology Based On Millimeter Wave Radar

With the rapid advancement of infrastructure construction in my country,the construction of expressways continues to extend from plains to mountains,and the proportion of bridges and tunnels is gradually increasing.However,these environmental and geographical conditions are complex,the space is closed and narrow,the longitudinal distance is long,and the shielding effect on signals is serious.It makes the traditional traffic monitoring method impossible to implement,and the millimeter-wave radar can better adapt to these environments and make up for the deficiencies of other sensors because of its high detection accuracy,small data processing capacity and strong adaptability to harsh climates.Therefore,in view of this background and demand,this paper conducts research on vehicle classification,recognition and tracking algorithm based on millimeter-wave radar.Firstly,the composition and algorithm support capability of AWR1642 millimeter-wave radar system are analyzed.According to the working principle and signal model of FMCW radar,the method of obtaining point cloud data such as target relative distance,radial velocity,angle and reflection intensity is studied.Secondly,according to the characteristics of millimeter wave radar point cloud,an improved A neighborhood DBSCAN clustering algorithm is studied to realize vehicle identification.According to the difference of the reflected point cloud at different distances from the vehicle to the radar,the radar detection scene is divided to achieve vehicle classification.Finally,by analyzing the correlation algorithm of point cloud data and the classical nonlinear Kalman filter algorithm in multi-object scenes,a vehicle tracking algorithm based on JPDA-EKF is designed for the highway tunnel scene,and its feasibility is verified by modeling and simulation.The algorithm studied in this paper is applied to the AWR1642 millimeter-wave radar system,and several field tests are carried out in the Jialiangshan expressway tunnel to analyze and compare the real situation of the actual scene and the observation results of the radar system.The classification,identification and tracking of vehicles has an accuracy rate of more than 90%.