Augmented unbiased minimum-variance input and state estimation for tracking a maneuvering satellite

Abstract

The technique of tracking a maneuvering satellite is significantly important for Space Situation Awareness (SSA). Traditional Kalman filters (KF) cannot robustly track unknown maneuvers. Motivated by the problem of tracking a non-cooperative maneuvering satellite, an augmented unbiased minimum-variance input and state estimation (AUMVISE) method is developed for estimating the state and the maneuver acceleration in this study. The maneuver acceleration estimate of the proposed method is proven to be more accurate than the original unbiased minimum-variance input and state estimation (UMVISE) method. Approaches based on the measurement residuals and the acceleration estimates are developed for maneuver start and end detection. The filter is switched between the AUMVISE method and the classical extended Kalman filter (EKF) to obtain an accurate tracking result during both the maneuver period and the non-maneuver period. Simulation results show that the proposed method has a suitable maneuver detection delay and outperforms the UMVISE method in estimating the state and maneuver acceleration.