Adaptive Disturbance Torque Estimation for Orbiting Spacecraft Using Recursive Least-Squares Methods

This paper develops a novel disturbance torque estimator for an orbiting spacecraft by using the adaptive least-squares parameter estimation technique. The disturbance estimation is first formulated as an adaptive least-squares minimization problem using a set of polynomial functions and then integrated with the feedback momentum estimator. The covariance update law with a variable forgetting factor is used, and it is shown that the convergent rate for estimation errors can be made at the same level as the forgetting factor. The proposed approach is particularly suited for orbiting small or microsatellite applications, where the momentum management capacity is often limited. The onboard estimated disturbance torque input can then be used as a part of control resource for spacecraft momentum management. The simulation results demonstrate the efficacy of the proposed concept.