Distributed Multi-Sensor Multi-Target Track Matching Algorithm Based on LMB Filter

Kuiwu Wang

doi:10.1016/j.dsp.2026.105953

2026-01-29

PRE

DOI：10.1016/j.dsp.2026.105953

原文链接

原文求助

摘要

En 中文

• In the multi-sensor fusion link, the existing method is to directly fuse the Gaussian components with larger weights to extract the state, which will lead to the error merging of adjacent targets and clutter interference. By comprehensively considering the distance measure and motion direction of the target, the proposed method realizes the combination of correlated Gaussian components and the pruning of uncorrelated Gaussian components, thus effectively fusing the data information representing the same target. This method not only improves the fusion accuracy of sensor networks, but also improves the multi-target number estimation accuracy of AA fusion. • In the multi-sensor tag matching phase, label history information is incorporated, and a label history similarity statistic is devised to provide a reliable criterion for assessing the quality of tag matches. Furthermore, tailored label matching methods are designed for different sensor network communication modes, and a genetic algorithm is introduced during the global label matching optimization stage to identify the optimal or suboptimal solution for global label matching. • In the multi-sensor tag matching phase, label history information is incorporated, and a label history similarity statistic is devised to provide a reliable criterion for assessing the quality of tag matches. Furthermore, tailored label matching methods are designed for different sensor network communication modes, and a genetic algorithm is introduced during the global label matching optimization stage to identify the optimal or suboptimal solution for global label matching.