Robust distributed autonomous orbit determination for LEO mega constellation based on inter-satellite ranging and astronomical measurement

Abstract

Low earth orbit (LEO) satellites are pivotal for the development of mega satellite constellations (MSC), offering low latency and global coverage. However, the rising number of satellites in LEO MSC presents significant challenges, particularly in terms of computational burden and data processing requirements. In this study, a robust distributed autonomous orbit determination (AOD) for LEO MSC based on inter-satellite link (ISL) ranging and astronomical observations is proposed. Specifically, first, we fully consider the advantages of ISL observations in autonomous navigation, and utilize astronomical observations to compensate for the missing datum mark problem. Second, to address the significant computational burden resulting from the large number of satellites in the LEO MSC, a distributed AOD strategy that combines active and auxiliary satellites is given. In addition, the robust AOD method of balanced extended kalman filter (BEKF) for LEO MSC is derived. The results show that the proposed distributed strategy with robust parameters estimation method can achieve convergence time within 30 min even with up to 1600 satellites, while the traditional method needs over 240 min. In addition, the combination of ISL ranging for distance constraints and astronomical observations for directional constraints significantly enhances accuracy in T and N directions by approximately 20 %. Furthermore, the error accumulation effect is reduced by approximately 64 % compared to traditional methods as the number of satellites increases from 400 to 1600. Moreover, as the complexity of LEO MSC increases, the traditional amplified measurement covariance extended kalman filter (AMCEKF) method may become ineffective. However, the robust BEKF method enables satellites with lower orbit accuracy to make incremental adjustments, while satellites with higher orbit accuracy receive fewer adjustments. Therefore, the proposed robust method of BEKF balances orbit accuracy among satellites, and the stability of the BEKF method is improved compared to the traditional method.