Computationally efficient distributed model predictive control of satellite swarms

This study develops a novel distributed model predictive control (DMPC) algorithm for the collision-free formation reconfiguration of satellite swarms. The multi-constrained optimization problem of large-scale systems is computationally heavy due to the variable coupling between satellites. To solve this, a new iterative algorithm is proposed to ensure the collision avoidance effectiveness of distributed optimization. In this iterative algorithm, the predicted trajectory is designed to decouple the state variables, and a novel collision avoidance rule is presented to guide the satellites to avoid each other reasonably. Furthermore, to achieve the swarm maneuver to arbitrary formation, a new terminal constraint is developed to track the time-varying desired state. The feasibility and stability of the DMPC algorithm under the influence of disturbance are demonstrated. Finally, the simulation results testify the efficacy of the developed algorithm, indicating that the algorithm is robust to disturbance and achieves collision avoidance while significantly reducing the computation time.