This paper presents innovative approaches for guidance of spacecraft formations during reconfiguration, developed under the paradigms of cluster autonomy and safe maneuverability. A task-assignment algorithm identifies those satellites that need to be rearranged in order to minimize the total propellant consumption. The decision process relies on convex optimization procedures which estimate control laws and maneuvering trajectories, accounting for thruster technology constraints and collision avoidance. The path-planning has a twofold implementation. In the centralized approach, the reconfiguration strategy is elaborated by a single unit: the maneuvering trajectories are generated considering solely collision avoidance constraints at expenses of computational costs. The alternative method is conceived such that each satellite autonomously determines its tasks. This de-centralized architecture allows the algorithm to run faster, but entails a higher traffic on the intersatellite communication link. In addition, collision avoidance is approximated to make each optimization independent from the other. For preliminary validation of the developed routines, a specific mission scenario is considered, with a distributed Synthetic Aperture Radar in a low Earth orbit that has to be reconfigured for tomographic applications. Simulation results show that, once remote-sensing requirements are specified within the limits set by standard products, the developed algorithms can manage autonomously the maneuver towards the corresponding operative pattern. In details, the centralized approach allows the reconfiguration to be handled with mean delta-v savings around 30%. On the contrary, strategies are elaborated by the decentralized planners about 7 times faster.

This paper presents innovative approaches for guidance of spacecraft formations during reconfiguration. The proposed strategies are developed under the paradigms of cluster autonomy and safe maneuverability. To that end, a task-assignment algorithm identifies the satellites that need to be rearranged in order to minimize the total propellant consumption. The decision process relies on convex optimization procedures aimed at estimating for each candidate satellite the control law and the maneuvering trajectory that satisfy constraints imposed by the thruster technology and collision avoidance. Two different versions of the path-planning algorithm are proposed. In the centralized approach, the reconfiguration strategy is elaborated by a single unit: the maneuvering trajectories are generated considering solely collision avoidance constraints at expenses of computational costs. The alternative method is conceived such that each satellite autonomously determines its tasks. This de-centralized architecture allows the algorithm to run faster, but entails a higher traffic on the intersatellite communication link. In addition, collision avoidance is approximated to render each optimization independent from the other. For the preliminary validation of the developed routines, a specific mission scenario is considered, with a distributed Synthetic Aperture Radar in a low Earth orbit that has to be reconfigured for tomographic applications. Simulation results show that, once remote-sensing requirements are specified within the limits set by standard products, the developed algorithms can manage autonomously the manoeuvre towards the corresponding operative pattern.