Dual-loop filtering algorithm of magnetometer measurements for attitude parameter estimation of geostationary spacecraft

Abstract

Effective operation of geostationary spacecraft implementing a wide range of space measurement services is contingent upon the high accuracy of their onboard attitude control system. Currently, the primary sensors for geostationary spacecraft attitude parameter determination are sun sensors, star trackers, and magnetometers. However, factors such as the high cost of sun and star sensors, their limited field of view (which hinders attitude determination during spacecraft rotation), and the high computational cost of processing star tracker data have spurred increased research into magnetometer-only attitude estimation methods capable of functioning even during highly dynamic rotation. Nevertheless, this introduces the challenge of achieving stable and accurate estimation amidst inevitable magnetic field measurement noise caused by various random physical factors. To date, various methods and algorithms have been proposed to mitigate magnetometer noise, with the Extended Kalman Filter being among the most effective. Major drawbacks of these methods include their reliance on specific cases of spacecraft angular motion and simplified noise models, necessitating additional Extended Kalman Filter tuning tailored to specific application scenarios. In this regard, this article proposes a new approach for synthesizing a stochastic attitude estimation algorithm for geostationary spacecrafts. The algorithm is invariant to the nature of angular motion and ensures stability and required estimation accuracy under general assumptions regarding magnetometer measurement noise. The proposed stochastic filtering algorithm incorporates an additional filtering loop that utilizes the Extended Kalman Filter output estimate vector as a linear observer for the angular parameter vector. This dual-loop filtering approach ensures both stability and estimation accuracy of angular parameters, fully meeting modern requirements for geostationary spacecraft onboard attitude systems. Numerical experiment results are presented to illustrate the effectiveness of the proposed approach.