A Huber-based robust tracking algorithm for orbital satellites

Abstract

In satellite tracking, atmospheric disturbances and sensor malfunctions frequently induce heavy-tailed distributions or outliers in the measurement noise, which consequently degrade the estimation accuracy of the standard converted measurement Kalman filter (CMKF) and may even lead to filter divergence. Although the CMKF transforms angular measurements into Cartesian coordinates to accommodate linear filtering frameworks, the resultant noise correlation and outlier contamination severely deteriorate the state estimation performance. To address this issue, this paper proposes a robust CMKF tracking algorithm based on the Huber methodology, designated as the Huber-based converted measurement Kalman filter (HCMKF). The key novelty lies in a component-wise residual weighting that adaptively corrects the measurement covariance matrix, enabling effective outlier suppression while retaining the correlated structure of the converted measurements. The proposed filter incorporates the Huber M-estimation into the measurement update stage and constructs a hybrid l1/l2-norm cost function, which adaptively suppresses the influence of measurements with large residuals. In comparison with the standard CMKF, the HCMKF exhibits significantly enhanced robustness and estimation accuracy under non-Gaussian noise and anomalous observation conditions, thereby providing an effective solution for engineering antenna servo tracking systems. The superior performance of the proposed algorithm is substantiated through simulation experimental validation.