多传感器多目标联合概率数据关联研究

Research on Joint Probabilistic Data Association Algorithm for Multisensor-multitarget Tracking

联合概率数据关联(JPDA)算法对单传感器多目标跟踪是一种良好的算法,但对于多传感器密集多目标跟踪,则计算量剧增,数据关联成功率下降。因此,改进联合概率数据关联(AJPDA)算法对多传感器多目标量测进行同源划分及单一传感器测量数据转换,然后采用JPDA算法求解空间目标轨迹交叉时的数据关联。仿真结果表明,AJPDA算法提高了成功关联概率,降低了求解数据关联概率的难度,可以解决密集目标的正确跟踪问题。

The joint probabilistic data association （JPDA） algorithm is a good method for the single sensor muhitarget tracking. However, for the muhisensor-muhitarget （MSMT） tracking in clutter, its calculation load comes higher and it may cause the incorrect association data. Therefore, the amended joint probabilistic data association （AJPDA） algorithm for MSMT tracking is proposed in this paper. The same source observations are classified into the same set. Then the JPDA algorithm can be used to obtain the data association when the space target traces crossing. The simulation results show that the AJPI）A algorithm can increase the successful rate of data association, and reduce calculation complexity. This algorithm can make the sensor correctly track densely distributed targets.