Data-Driven Optimal Formation-Containment Control for A Group of Spacecrafts Subject to Switching Topologies

This research focuses on the data-driven optimal time-varying formation-containment control problem for a group of heterogeneous spacecrafts that are simultaneously subject to switching topologies, nonlinear relative motion dynamics, and parameter variations. A distributed observer is developed for each spacecraft to generate the position references and to decrease the influences of switching topologies. The optimal formation-containment control protocol is learned using relative motion data and system inputs, based on the reinforcement learning theory. Simulations of a spacecraft group are given to show the performance of the proposed formation-containment control protocol.