SPACECRAFT SWARM RECONFIGURATION BASED ON CENTRALIZED PLANNING AND ARTIFICIAL POTENTIAL FIELD FEEDBACK

This paper investigates the optimal reconfiguration of spacecraft swarm. The traditional spacecraft formation specifies the target point with each spacecraft and then converts the reconfiguration problem into a path planning problem. As the number of the swarm increases, a more optimal solution can be obtained by matching the target point with the spacecraft one by one. In this paper, the geometries of the expected configuration are specified without specifying the expected locations of individual spacecraft. The first step is to select an optimal match to minimize the maneuver time of the reconfiguration mission. This paper uses the genetic algorithm to find the optimal match, with the time taken as the target function. Next, a Lyapunov based nonlinear controller with the awareness of collision avoidance is developed to accomplish the reconfiguration process. Owing to the reconstruction process running in a geostationary orbital environment, the dynamic equations are established in the Hill coordinate system. The reconfiguration process is a Rest-to-Rest process. The Lyapunov function is defined using the condition that each spacecraft reaches the target point with zero speed. To achieve avoidance of potential collision with fixed obstacles and neighboring spacecraft, the Lyapunov's method is extended to artificial potential functions, which generates repulsive forces from the obstacles and neighboring spacecraft. With the negative derivative of the Lyapunov energy function, the "repulsion" force of the artificial potential functions and the angular velocity produced by gravity, nonlinear controller are set up to accomplish the path planning of the reconfiguration mission. The numerical simulations show that spacecraft swarm reconfiguration based on time-optimal centralized planning and artificial potential field feedback is effective.