Resolvable Group State Estimation with Maneuver Movement Based on Labeled RFS

Abstract

This paper considers the problem of tracking multiple resolvable group targets using the labeled random finite set framework. While the generalized labeled multi-Bernoulli (GLMB) filter is an efficient multi-target tracking filter, it cannot capture the the dependence or correlation between members of each group. In this paper, we introduce a group target model by incorporating graph theory into the labeled random finite set framework, which accounts for dependence between group members. We then propose a GLMB approximation of the prediction and update step of the Bayes filter for multiple resolvable group targets. Simulation are presented to benchmark the proposed filter against the GLMB filter.