基于Keystone变换与Dechirping补偿的雷达运动目标检测跟踪算法研究

Research on Radar Moving Target Detection and Tracking Algorithm Based on Keystone Transform and Dechirping Compensation

对目标回波信号进行长时间积累,是在不对雷达硬件系统进行升级改造的情况下,提高回波信噪比及雷达系统探测能力的低成本实现方法。本文主要对目标沿雷达径向匀加速运动时导致的距离徙动和多普勒走动问题进行研究,提出了基于Keystone变换和Dechirping补偿的积累算法。在没有先验信息的情况下通过Keystone变换实现包络对齐,而后利用对加速度进行搜索补偿,来补偿多普勒走动以提高雷达积累增益。最后基于AR模型的卡尔曼滤波算法实现机动目标的跟踪。利用空天杯比赛数据可以证明算法在检测微弱运动目标方面的优势。

The long time integration method of radar target echo signal is a low-cost method to improve the SNR ratio and the detection capability of radar system without upgrading the radar hardware system. In this paper, the across range unit and Doppler frequence migration caused by the uniform acceleration of the target along the radial direction of radar are studied, and an integration algorithm based on Keystone transform and Dechirping compensation is proposed. In the absence of prior information, the Keystone transform is adopted to realize envelope alignment, and then the searching compensation for acceleration is performed to compensate Doppler frequency migration and to improve the radar accumulated gain. Finally, Kalman filter algorithm based on AR model is used to track maneuvering targets. It can be proved that the algorithm can detect the weak moving target by using the real data.