Research On Target Detection Algorithm Based On Millimeter Wave Radar

The millimeter wave radar installed on the vehicle is an effective way to solve the problem of the increasing traffic accidents.The domestic research on 77GHz vehicle millimeter wave radar is still at the stage of theoretical research and experimental study,which is not advanced compared with European and American countries.The target detection algorithm is one of the key technologies for the research of millimeter wave radar,and it is of great significance for its intensive study.In this paper,the detection algorithm of multi-targets distance,speed,angle is studied based on the saw-tooth linear frequency modulation continuous wave system in millimeter wave radar,and the algorithm is simulated and verified on MATLAB and radar simulator.The main research work of this article is as followings:(1)Firstly,the background,development status and future development trend of automobile millimeter wave radar and target detection algorithm are illustrated.The saw-tooth wave chirp continuous wave is selected as the transmitted signal,and the speed ambiguity and distance speed coupling problems in the saw-tooth wave system are analyzed.The waveform parameters are designed according to the distance range,speed range,range resolution,angle resolution,etc.of the millimeter wave radar and the process of target detection algorithm is built.(2)Secondly,the module in the process of target detection algorithm is optimized.Aiming at the problem of speed ambiguity and distance-velocity coupling,a new algorithm is proposed which makes use of 128 chirps and 16 chirps of periodic T1 and T2 respectively to solve the speed ambiguity,and then corrects the target distance;The single-target and multi-target detection performance of the unit average,ordered and feedback constant false alarm detectors under the background of uniform clutter are compared,and the feedback constant false alarm detector with better performance is used to filter the signal;Finally,Texas Instruments'AWR1243 evaluation board and DCA1000 acquisition card are used to collect target data in the radar simulator to verify the feasibility of the algorithm.It can realize the medium-distance detection of 80m,the range resolution is 0.207m,the speed resolution is 0.1572m/s,the distance measuring error is within 0.5m,and the speed measuring error is within 1.5m/s.(3)Finally,the angle estimation algorithm based on FFT and MUSIC and the DO A estimation algorithm of coprime array are studied.First,the detected target speed is used to compensate the phase error of the MIMO radar,and the DOA estimation is combined with the FFT and MUSIC algorithm to reduce the spectral peak search interval to reduce the calculation amount;Then the expansion method of coprime array and the covariance matrix reconstruction DOA estimation algorithm under the array are studied.This algorithm makes use of non-uniform virtual array of coprime array to constructs a uniform linear array by introducing the idea of virtual array interpolation.After constructing the covariance matrix of a uniform linear array,the DOA estimation is performed,and the spatial spectrum,resolution,and the number of snapshots,the number of incident signal sources,and the different signal-to-noise ratios on the root mean square error of the algorithm are analyzed in detail.The variance matrix reconstruction DOA estimation algorithm can achieve DOA estimation of 9 source numbers when the number of elements is 7,it can distinguish the adjacent target with the incident angle?1=-1.5°,?2=1.5° and has good resolution of the angle of arrival.