论文部分内容阅读
在分布式多目标跟踪系统中,由于局部传感器开机时间、采样频率以及通信延迟不同等原因,导致来自各传感器的局部航迹往往是异步不等速率的。目前一般的方法是先进行时域配准再进行航迹关联,但是在同步化的过程中,航迹估计值的误差会发生传播,影响航迹关联的性能。针对此问题,提出了一种基于区实混合序列相似度的异步不等速率航迹关联算法。算法首先通过区间数-实数混合序列变换(IRST)得到等长度的航迹行为序列,然后定义一种新的序列差异信息度量,得到混合序列的相似度,以此进行航迹关联判定。仿真实验表明,该算法可以有效地解决异步不等速率航迹关联问题,并且通信延迟和数据乱序对算法性能的影响不明显。
In distributed multi-target tracking system, the local track from each sensor tends to be asynchronous and unequal due to the different start-up time, sampling frequency and communication delay of local sensors. At present, the common method is to perform time-domain registration and track association first, but during the process of synchronization, the error of track estimation will propagate and affect the performance of track association. In order to solve this problem, an asynchronous unequal-rate trajectory association algorithm based on real-world mixed sequence similarity is proposed. The algorithm first obtains the equal-length trajectory sequence through interval number-real mixed sequence transformation (IRST), and then defines a new sequence difference information metric to obtain the similarity of the hybrid sequence to make the trajectory association decision. The simulation results show that this algorithm can effectively solve the problem of asynchronous unequal-speed track association, and the communication delay and data disorder do not have obvious influence on the performance of the algorithm.