基于GMPHD的雷达组网检测跟踪算法研究

Study on Track and Detect Algorithm in Radar Networking Based on GMPHD

将高斯混合概率假设密度算法(Gaussian mixture probability hypothesis density algorithm,GMPHDA)成功应用于多雷达组网跟踪检测弱信噪比多目标,能估计得到所有目标状态与数量,但其跟踪结果是估计值随机集,未与各真实目标分别对应,目前未出现相关完整算法。因此提出对估计航迹进行辨识,包括航迹区分、继续、新生与恢复,给出了一整套航迹辨识算法流程,完善了多雷达组网跟踪检测目标算法。仿真结果表明,能跟踪检测到弱信噪比环境下所有目标,提出的航迹辨识算法能够形成与各真实目标一一对应、逼近的航迹。

The states and number of WSNR multi-target are tracked accurately by Gaussian mixture probability hypothesis density algorithm （GMPHDA） application. But tracking result is random set of target states, it doesn＇t correspond one to one with real targets. And the complete algorithm about corresponding has not been proposed. To get target tracks corresponding with real targets, a suite of algorithm about identifying target tracks is proposed, called track identification algorithm, which contains track distinction, continuance, newborn and restoration. The track identification algorithm improves track and detect algorithm in multi-radar networking. Simulation results show that WSNR multi-target is tracked in multi-radar networking, which gets target tracks corresponding one to one with real targets by the proposed identification algorithm.