Orbit-Attitude Configuration-Keeping for Mega-Constellation Using Collective Iterative Learning Control

To decrease the self-induced collision risk and keep an identical view or antenna pointing conditions, the configuration of the mega-constellation, including the orbital position and the attitude, needs to be consistent with high precision. Specifically for the same orbital plane, the spacecraft are subject to the homogeneous periodic perturbation from the irregular gravitational field and track the identical reference trajectory and attitude. A scheme of collective iterative learning control (CILC) using inter-spacecraft exchange information is presented to consider both the fast convergence rate and small tracking error. The six-degree-of-freedom kinematics and dynamics are formulated on the Lie group SE(3), avoiding the nonuniqueness and unwinding phenomenon in attitude control compared with the dual quaternions. The proposed CILC is illustrated in detail through multi-spacecraft configuration-keeping scenarios and Monte Carlo analysis, demonstrating enhanced robustness to space environment uncertainties and reduced effort in parameter adjustment.