一种基于随机集的PHD多目标多传感器关联算法

针对复杂电磁环境下的多目标关联计算量大、准确率低的问题,提出了一种基于随机集概率假设密度(PHD)的多目标多传感器关联算法。该方法首先采用高斯混合PHD(GMPHD)对多传感器的量测信息进行滤波,再对滤波结果做最近邻数据关联处理,从而得到多目标航迹。杂波环境下的仿真实验表明,该方法在保证滤波精度的同时,能够有效降低运算量,提高数据关联的准确度。

A PHD Algorithm of Multi-Target Multi-Sensor Association Based on Random Finite Set

Large computation cost and low accuracy of data association are for classical associa- tion algorithms under the condition of complicated multiple targets. Aiming at practical applica- tion, multi-target multi-sensor association algorithm based on random finite set PHD is pro- posed. The information of multi-sensor radiation source with PHD is filtered and the treated multi-target data is associated. Simulation results demonstrate the computational cost is de- creased and the accuracy of data association is improved by the proposed algorithm with the guar- antee of filtering accuracy, which is fit for multi-target muhi-sensor association and recognition in complicated environment.