Tube-based model predictive control for the approach maneuver of a spacecraft to a free-tumbling target satellite

Abstract

Rendezvous and proximity maneuvers are historically critical operations. Such maneuvers include the approach and docking of a transport vessel to the International Space Station or the recovery of a tumbling satellite. Numerous control approaches have been proposed in recent decades for handling such maneuvers. In the past decade, many of these approaches have been based in Model Predictive Control. In this work, a tube-based robust model predictive controller is proposed for the robust control of a rendezvous maneuver. In the first part of this thesis, the theory for rendezvous and proximity maneuvers and for nominal and tube-based robust model predictive control is presented. Tube-based robust model predictive control is then applied to the well-studied double pendulum problem. The design methodology and controller are implemented in software, and the controller is simulated to reproduce the results presented for the application of this control method to the double pendulum problem in literature. In the second part of this thesis, the tube-based robust model predictive control framework is applied to a rendezvous maneuver with the unresponsive, tumbling Envisat. Uncertainty is introduced to this application of robust model predictive control through the uncertainty in the motion of the Envisat satellite. The chaser spacecraft will be required to successfully rendezvous with Envisat while tracking a pre-determined optimal trajectory and robustly satisfying the system constraints.