Local Information based Organization of Distributed Spacecraft Swarm Using Artificial Potential Field

This paper investigates the problem of distributed organization of spacecraft swarm where each spacecraft is an independent agent that can interact with neighbouring members within a certain distance. According to the global objective of the organization, each agent makes decisions on its behaviour based on its own status and adjacent members within a certain distance. By a carefully designed strategy and under certain conditions, the swarm can accomplish the global objective. In the configuration reconstruction mission, only the expected global configuration is specified. The target position of each spacecraft is uncertain and not fixed. At the beginning of the reconstruction, every spacecraft can only determine the initial target point according to certain rules. It is possible to have conflicts of destination point selections since all spacecraft can only get local information. Each spacecraft constantly adjust their target point according to the local information in the process of accomplishing this task, until the group converges into the specified global configurations. In this paper, three different target allocation strategies are designed, so that each spacecraft can do continuous selection and optimization, and finally determine a non-conflict target matching to complete the formation reconstruction task. The artificial potential field method is utilized to control each spacecraft and complete the path planning, while ensuring that the spacecraft swarm to achieve collision avoidance and other requirements. In addition, the time and fuel consumption are compared. The numerical simulations show that these three strategies make obvious optimization effects.