Implementation And Research Of Tracking Fusion And Target Tracking Algorithm For Defensive Millimeter Wave Radar

Millimeter wave radar has its outstanding advantages in the field of defense,with its advantages of long detection range,all-weather working all day,strong antiinterference ability,real-time provision of target azimuth attribute information,simple erection and low maintenance cost,has become a hot spot in the field of security research at home and abroad.Although the theory of target tracking has been relatively mature,the application of millimeter wave radar in the scene of important defense has not been long.There is still a lot of research space,such as choose the data association filtering algorithm and how to improve the algorithm performance in combination with the specific application environment.In this paper,the target tracking technology of millimeter wave radar under defensive scene is studied,which mainly includes the following points:1.According to the test data in the practical application environment,the filtering effect of Kalman filtering algorithm and extended Kalman filtering are compared and verified.After determining the application of Kalman filter,combined with the actual data,the reason why the far-end filtering effect is poor is analyzed,and the conclusion that the low-echo signal-to-noise ratio affects the filtering effect by affecting the accuracy of the angle measurement is obtained.Based on this conclusion,a Kalman filter algorithm with angle and signal-to-noise ratio weighting is proposed,which effectively reduces the fluctuation of the filter trajectory in the distance.2.For the problem of single target tracking,based on the collected linear moving target data and nonlinear moving target data,the tracking performance of the nearest neighbor association algorithm and the probability data association algorithm is compared and analyzed,and the tracking performance of the two algorithms under different degrees of clutter environment is given.Aiming at the problem of multi-target tracking,the performance differences of the joint probability data association algorithm in the three aspects of tracking accuracy,algorithm time-consuming and false tracking rate and the empirical joint probability data association algorithm are verified.3.Aiming at the problem of extended target tracking encountered in practical product applications,an extended target tracking method based on cosine theorem is proposed,which can be used for extended target tracking under single target and sparse multi-target conditions.Then,an extended target tracking method suitable for high resolution radar is studied and analyzed,that is,the extended target tracking algorithm based on shape model.Compared with the tracking method without this processing,the tracking accuracy can be effectively improved.