Fast preliminary design of low-thrust trajectories for multi-asteroid exploration

Multiple-asteroid exploration with low-thrust propulsion requires the design of transfer trajectory and the selection of the visiting sequence, as well as the estimation of the propellant budget. To this end, this paper presents a method to generate transfer trajectories and visiting sequences rapidly using finite Fourier series (FFS) and Monte-Carlo Tree Search (MCTS). The FFS method can generate the transfer trajectory rapidly, which can provide suitable initial approximations, leading to more accurate trajectory optimizations. This study adopts the MCTS algorithm for asteroid sequence selection in the multi-asteroid exploration. By comparing with the traversal algorithm, the greedy algorithm and the tree search algorithm with the trimming strategy, the numerical results show that the MCTS can be used to obtain a quasi-optimal sequence with higher probability and less computation time. Consequently, a method combining FFS and MCTS can rapidly acquire the quasi-optimal visiting sequences with a large probability and the suitable initial trajectory for a multi-asteroid exploration mission. This is very important for the rapid feasibility assessment of hundreds of flight scenarios at the preliminary mission design stage. (C) 2019 Elsevier Masson SAS. All rights reserved.