Optimal scheduling of multiple Sun-synchronous orbit satellites refueling

The scheduling problem of numerous Sun-synchronous orbit satellites refueling with multiple resupply spacecraft (RSc) considering J2 perturbation is raised in this paper. Problem scenario indicates that a large number of Sun-synchronous orbit satellites require refueling of different weight, depending on their fuel surplus. We propose a new strategy for the refueling problem, in which the refueling campaign is composed of several independent missions and each mission is carried out by a single RSc. The RSc will transfer between objective satellites so as to refuel them in order. The optimal problem of this strategy can be classified as a Multi-person Traveling Salesman Problem (MTSP). Hence we present a two-step optimization model to solve it, using refueling sequences as design variables and the estimate cost as the objective function. Searching for an initial solution is the first step of optimization, where a database is generated and Branch-and-bound algorithm, Ant Colony System (ACS) are deployed. The second step is to perform local optimization with Simulated Annealing (SA). At last, we give four different numerical examples to evaluate the effectiveness and robustness of this model under different initial conditions, and discuss the impact of these conditions on the optimal results.