Input matrix compensated strong tracking filter for maneuvering spacecraft tracking

Abstract

Accurate tracking of maneuvering space targets is crucial, as their unpredictable movements pose significant challenges in increasingly congested space environments. A widely used method for tracking impulsive maneuvering targets is the Strong Tracking Filter (STF). While effective, this method has two primary limitations in practical applications. First, it causes significant transient tracking errors following a maneuver. Second, because the method relies on measurement residuals to detect maneuvers, it is unable to distinguish between actual maneuvers and measurement outliers, leading to misinterpretations that degrade tracking accuracy. This paper proposes two key improvements to overcome these issues. First, an input matrix compensation framework is introduced based on a residual orthogonalization criterion, which updates the covariance in a way that better reflects the physical impact of unknown maneuvers through the system’s input matrix. This modification effectively eliminates transient tracking error overshoots while maintaining tracking accuracy. Second, an auxiliary filter is introduced to handle measurement outliers, allowing for precise differentiation between outliers and maneuvers, thereby enhancing the algorithm’s robustness in the presence of outliers. Simulation results demonstrate that the proposed method outperforms the STF in terms of convergence, accuracy, and robustness, particularly in scenarios with measurement outliers.