ESA GNC Conference Papers Repository

Autonomous and magnetometer-based orbit determination algorithm in target pointing satellite for speeding up filter convergence is proposed. The cosine angle between geomagnetic field vector and target direction vector, as viewed from the satellite, provides some quantity that is sensitive to satellite position, more regular and periodical than the magnetic field magnitude as a satellite moves around its orbit, and easily observable from magnetometers. An algorithm is formulated by applying the unscented Kalman filter (UKF) based on the laws of orbital mechanics, the measurement models of geomagnetic field magnitude and the cosine angle. The algorithm is simulation tested and verified using real-flight data from the Magnetic Field Satellite (MAGSAT). The results obtained from numerical simulations and MAGSAT data illustrate a significant convergence speed up, moderate position accuracy and robustness against both noise and bias of sensors. The proposed orbit determination is adequate as a primary routine for a target pointing satellite mission that requires fast convergence and moderate position accuracy. Otherwise, it is sufficient as a backup system to provide redundancy.