This chapter applies the concepts presented in the previous ones to design a MPC for generating orbital rendezvous trajectories on board the chaser spacecraft, considering elliptical target orbits. We first present a naive formulation with a state quadratic Lagrangian cost and the receding-horizon strategy and demonstrate through simulation that it is not suitable for this application. This is iteratively improved until we reach two fuel-optimal formulations, known as Finite-Horizon MPC and Variable-Horizon MPCs, where the former has fixed final time and is formulated as a LP, and the latter has free final time and is formulated as a MILP. The passive safety constraint is approached, which makes the problem non-convex, and a new method for efficiently solving it via sequential linear programming is presented. We also cover the topic of robustness to disturbances and perturbations and discuss different techniques for granting robustness, including two contributions in the present work. Finally, we perform several simulations demonstrating the above methods, including in a high-fidelity industrial simulator, with emphasis on illustrating that they may be feasibly implemented in real-time and on-board spacecraft.
Rendezvous with Model Predictive Control
SpringerBriefs in Applied Sciences
2021-06-25
49 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Predictive control for spacecraft rendezvous
TIBKAT | 2021
|Distributionally Model Predictive Control for Spacecraft Rendezvous and Docking
Springer Verlag | 2021
|