Robust model predictive control for automated rendezvous maneuvers in near-earth and moon proximity

Automated rendezvous and docking has been identified by the space community as one of the cornerstone technologies to enable and support the new era of space exploration towards the deep space. The strict safety requirements and the limited actuation capabilities require high-reliable controllability during the close-range rendezvous, i.e., closing and final approach phases. The presence of persistent environmental disturbances that affect the spacecraft trajectory and attitude can compromise the mission success. This work presents a tube-based robust model predictive control approach for a dual-environment spacecraft rendezvous problem in highly elliptic orbits. During the entire mission, the spacecraft is called to perform several rendezvous maneuvers in highly elliptic orbits, i.e., a Geostationary Transfer Orbit and a Near Rectilinear Halo Orbit. The same robust control is exploited in all rendezvous operations and the robustness of the controller is demonstrated under the presence of persistent internal and external disturbances. Moreover, the tube-based robust model predictive control performance have been compared with those of a classic linear-quadratic model predictive control, highlighting the effectiveness of the robust approach in the presence of disturbance, both in Earth and Moon proximity.