基于标签多贝努利多传感器组网目标跟踪算法

Multi-sensor networking target tracking algorithm based on labelled multi-Bernoulli

随着电磁环境的日益复杂,强干扰和高杂波带来的目标低检测概率问题日益突出,给探测系统准确估计监测区域内目标个数以及目标状态带来了新的挑战。针对低检测概率问题,提出随机有限集框架下基于标签多贝努利(labelled multi-Bernoulli,LMB)多传感器组网目标跟踪算法。该算法首次将LMB框架应用到不同探测范围的多传感器组网目标跟踪场景中,实现了多目标跟踪目标数和相应状态稳定估计。仿真结果表明,所提方法不仅能在低检测概率条件下获得目标稳定的航迹估计,及时捕捉目标新生、消亡等事件,还能有效叠加不同传感器不同探测范围,充分发挥多传感器优势。

With the increasing complexity of electromagnetic environment,strong interferences and high clutter rate lead to the low detection probability problem,which brings a new challenge for the observation system to accurately estimate the number and states of target in surveillance area.Aiming at the problem of low detection probability,a multi-sensor networking target tracking algorithm based on labelled multi-Bernoulli(LMB)in the framework of random finite sets is proposed.The proposed algorithm incorporates the LMB framework into target tracking scene of multi-sensor networking with different detection ranges for the first time and realizes the target number and corresponding sate stability estimation of multi-target tracking.The simulation results show that the proposed algorithm not only provides stable estimation of trajectories under the condition of low detection probability,including targets born and die,but also gives an effective superposition of different detection ranges,which makes full advantage of multiple sensors.