主要运用检测前跟踪动态规划(DP-TBD)算法解决不同分辨率的多传感器多目标融合问题,提出了一种基于DP-TBD的不同分辨率的传感器集中式异步融合算法。该算法分为两步:第一步将不同分辨率的传感器取得的回波数据统一对齐到相同单元格尺寸的量测空间;第二步将具有不同采样间隔的传感器得到的量测数据按照时间顺序传送至融合中心进行集中式融合。接着,用动态规划检测前跟踪算法来处理所得到的数据,并且进行航迹恢复。同时,为了提高异步融合的有效性,文中还提出了一种基于变转移状态数的算法。仿真结果说明基于DP-TBD的集中式异步融合算法的有效性,并且和单个传感器的性能比较,该算法大大改善了目标跟踪的性能。

This paper addresses multiple targets tracking problem using multiple different resolution sensors via dynamic programming (DP) based track-before-detect (TBD). To alleviate this problem, a centralized asynchronous fusion algorithm for sensors with variable resolution via DP-TBD is proposed in this paper. This algorithm is divided into two steps. In the first step, the echo data obtained from different resolution sensors are aligned to the measurement space of same cell size. In the second step, measurement data derived from sensors with different sampling periods is sent to the fusion center in time order. Next, DP-TBD is used to dispose these data and generate recover track. Besides, a method of adaptively adjusting transition states number is proposed to improve the efficiency of asynchronous fusion. Simulation results show that the proposed algorithm is able to solve sensors with different resolution a synchronous fusion problems and effectively improve the tracking performance of DP-TBD comparing to single sensor.

TN958

长江学者奖励计划;中央高校基本科研基金资助项目(ZYGX2016J031);中国博士后科学基金面上基金资助项目(2014M550465) 和特别基金资助项目(2016T90845)