基于S-D分配的多传感器联合概率数据互联算法

Multisensor joint probabilistic data association algorithm based on S-D assignments

为解决集中式多传感器系统中多目标跟踪问题,提出了一种基于S-D分配的集中式多传感器联合概率数据互联算法。算法首先应用广义S-D分配的规则对每个传感器送来的观测数据进行排列组合,然后对每个组合中各量测点进行概率加权以获得一个等效量测点,最后根据每个等效量测点的联合似然函数计算其联合互联概率并获得融合中心的状态估计。该文最后给出了该算法与已有集中式多传感器联合概率数据互联算法的仿真比较,仿真结果表明该文算法的跟踪性能更优越。

Multitarget tracking with centralized multiple sensors is improved using a multisensor joint probabilistic data association algorithm based on S-D assignments. The algorithm permutes and combines measurements from each sensor using the rules of the generalized S-D assignment algorithm. Then, all the measurements in each assignment are combined into one equivalent measurement according to their weights in the assignment. Finally, the joint association probability of each equivalent measurement is calculated to obtain the state estimate of the fusion center. Typical simulations show that the algorithm is more effective than the centralized multisensor joint probabilistic data association method.