GTOC12: Methods and results from the National University of Defense Technology

Abstract

This paper presents the solutions and results of the 12th edition of the Global Trajectory Optimization Competition (GTOC12) of the National University of Defense and Technology. To address the complex interstellar mining problem proposed by GTOC12, our solution is divided into two stages. The first stage focuses on preliminary work, including the target selection, the establishment of departure and return databases, and the development of methods to estimate transfer costs, with the aim of enhancing planning efficiency during the global planning phase. The second stage involves trajectory optimization for multiple mining ships, including single-mining-ship trajectory optimization and a multiship iterative process. For single-mining-ship trajectory optimization, the method involves three steps: first, employ a heuristic method for planning the first rendezvous sequences; second, utilize an ant colony optimization (ACO) algorithm for planning the second rendezvous sequences; and third, apply a differential evolution (DE) algorithm alongside an indirect method to refine rendezvous times and low-thrust trajectories. Through the implementation of a multiship iterative strategy, the team accomplished trajectory optimization for multiple mining ships that met the constraints. The final score submitted by the team was 15,160.946, which achieved the sixth place in the competition.