Robust Multiple Space Object Tracking Using the Possibility Generalized Labeled Multi-Bernoulli Filter

doi:10.34133/space.0263

Robust Multiple Space Object Tracking Using the Possibility Generalized Labeled Multi-Bernoulli Filter

2025-09-05

0

OA

AI

DOI：10.34133/space.0263

Original

Original request for help

Save

Abstract

En 中文

A major challenge in many multiple space object tracking algorithms lies in the inability to determine a complete probabilistic characterization of the different aspects of the system such as dynamics and observations. In this paper, a robust multiple space object tracking method is developed by leveraging the framework of outer probability measures. The possibility generalized labeled multi-Bernoulli (GLMB) filter is employed to handle epistemic uncertainty in the process of multiple space object tracking, such as the presence of ignorance in dynamical and observation models. In order to initiate the orbital state of birth targets in the context of the possibility GLMB filter, the possibilistic admissible region (PAR+) method is introduced to offer a reliable initial orbit determination based on imperfect information. The developed possibilistic PAR+GLMB scheme provides improved robustness in the case of partial knowledge about the system. The features of the proposed PAR+GLMB method are illustrated using 4 simulated space object tracking case studies.

Robust Multiple Space Object Tracking Using the Possibility Generalized Labeled Multi-Bernoulli Filter

Abstract

Journal

Organization