DECENTRALIZED, LOW-COMMUNICATION STATE ESTIMATION AND OPTIMAL GUIDANCE OF FORMATION FLYING SPACECRAFT

摘要

This paper presents an integrated approach to GNC of formation flying spacecraft. The Navigation algorithm estimating the full relative state of all the spacecraft is a full-order decentralized filter, based on an Extended Kalman Filter for local measurements, and on Covariance Intersection for the fusion between local state estimates and estimates communicated by other spacecraft, eliminating EKF divergence problems. For Guidance and Control, an algebraic closed-loop algorithm, based on Pontryagin's maximum principle, is proposed, minimizing the propellant consumption and ensuring collision avoidance. This algorithm is regularly recomputed. Simulation results for a GTO 3-spacecraft formation are presented.