Design And Implementation Of Target Tracking Algorithm For Vehicle Collision Avoidance Radar

Due to the continuous update of modern science and technology and the continued deepening of scholars,car anti-collision radars are now available from the first pure detection of obstacles around the car to ensure the driving safety of the driver.Has undergone swaying changes,up to the advanced driving support system of.(ADAS),providing drivers with a more comfortable and convenient driving experience.After receiving the target echo signal,the car collision prevention radar performs radar signal processing through the beat signal obtained by mixing the transmission signal and the transmission signal,and acquires point trace information on the target motion state parameters.Performs tracking filtering on the point trace data of the target.After processing,the estimated parameters of the target are obtained and finally the tracking trajectory is formed.In today's changing social road scenes,high requirements are placed on tracking performance,so research on target tracking algorithms becomes necessary.Under the background of automobile anti-collision radar,the thesis does the following work:1.The chip framework of the millimeter-wave automotive collision avoidance radar is analyzed.The raw ADC data about a single target moving object is collected through field experiments,and the radar signal processing is performed on the MATLAB digital platform to obtain the point trace information of the target object.The implementation of the target tracking algorithm provides an input source.The process of radar signal processing is mainly: two-dimensional fast Fourier transform,twodimensional constant false alarm detection,point condensation processing and target angle measurement algorithm processing.2.The linear Kalman filtering algorithm and the nonlinear Kalman filtering algorithm are analyzed in detail.The uniform velocity motion model,uniform acceleration motion model and turning model state equation and linear radar measurement equation in the measured data are established,and the linear Kalman filtering algorithm and the nonlinear Kalman filtering algorithm are implemented respectively.Target tracking with filtering algorithm.The experimental results show that the output after tracking filtering has smaller errors than the results measured by the observation radar,which meets the expectations of tracking filtering.3.Aiming at the situation that the moving trajectory changes during the tracking of moving targets,an interactive multi-model algorithm based on the Kalman filter algorithm without trace is implemented.Through simulation analysis,it is concluded that the interactive multi-model algorithm can still maintain a good tracking effect even when the motion model is not single.