基于证据理论的联合概率数据关联算法

Joint probability data association algorithm based on evidence theory

数据关联是目标跟踪技术中的核心部分,多目标情况下的数据关联技术更是研究的重点,由于多目标量测之间的互相干扰、外部环境干扰以及传感器性能等客观因素的约束,使得量测信息部分存在着相应的量测误差,密集环境中的多目标跟踪比较困难。针对这个问题,提出的新算法利用联合概率数据关联方法进行密集杂波环境下的数据关联,结合证据理论的思想对多传感器量测信息进行优化组合,有效地减小了量测误差对跟踪目标的影响。通过仿真结果可以看出,改进算法大大提高了跟踪精度,并具有良好的抗干扰能力,适用于解决工程实际问题。

Data association technology is the key part in multi-sensor target tracking systems, and is even more important under the circumstance of multitargets. Because of the measurements of multi-targets interfering each other, the lack of priori knowledge of tracking environment and restriction of sensor’s performance, the introduced error is unavoidable during the measuring process and the tracking is difficult. Aiming at solving these problems, a new algorithm based on the joint probability data association method combining with evidence theory is used to make association under a dense clutter environment. After optimization of multi-sensor information, the influence from measure errors is lowered. From the simulation result, it can be seen that the improved algorithm greatly advances tracking accurancy and has a favourable anti-jamming ability, which is suitable for dealing with engineering problems in practice.