被动传感器阵列中基于视线距离的数据关联

Data association based on line-of-sight range in passive sensor array

针对低检测概率环境下被动传感器阵列中的数据关联问题,提出了一种基于视线距离的数据关联新算法。首先详细分析了传统基于视线距离数据关联方法的不足,提出各个观测站对于同一个目标的平均视线距离应最小。同时,认为同一个目标只要被超过2/3的传感器所观测,此关联组合就有效。最后,利用关联组合的平均视线距离和有效传感器个数来设计模糊隶属度函数,以模糊关联隶属度最大为关联准则构建数据关联算法,给出了算法流程。实验结果表明,在检测概率不为1、误差较大的多目标环境下,提出的算法要明显优于传统方法。

Aiming at the data association problem of multiple target tracking in passive sensor array, a novel data association algorithm based on line-of-sight range is proposed for the low probability of target detection environment. Firstly, the disadvantages of the conventional line-of-sight range data association method are ana- lyzed. In order to deal with these disadvantages, the criterion that the average of line-of-sight range of the same target may be the shortest is proposed. At the same time, when the number of measurements of the same target exceeds 2/3 of sensor stations, the association is regarded as valid. Secondly, the fuzzy membership function based on the average of line-of-sight range and the number of valid sensors is designed. Finally, the data association algorithm is built. The simulation results show that the proposed algorithm is more accurate than the existing one for multiple target environments.