Neural-network-assisted Optimization of a Close-Range Multi-Spacecraft Rendezvous Mission Based on a Multi-Impulse Maneuvering Strategy

Huida Yan,Yanning Guo,Chuanjiang Li,Bin Song
DOI: https://doi.org/10.1016/j.asr.2023.04.043
IF: 2.611
2023-01-01
Advances in Space Research
Abstract:Rendezvous mission planning between service spacecraft and multiple close-range target spacecraft in the on-orbit service application is studied, using multiple impulses based on the Clohessy-Wiltshire equation as a maneuvering strategy. An impulsive correction method is proposed to avoid rendezvous errors. The optimization objectives of this work include the minimum total velocity increase and the minimum mission time. A two-level optimization approach is developed to find the Pareto-optimal set of the above multi-objective optimization problem. In the low-level optimization, the propellant-optimal impulsive trajectory for the rendezvous between the SSc and each target spacecraft at a fixed maneuver duration is solved by a proposed iterative optimization algorithm with a double loop. On the other hand, the rendezvous sequence and maneuver duration are optimized by a proposed neural-network-assisted evolutionary algorithm in up-level optimization, which combines the feedforward neural network with the elitist multi-objective genetic algorithm (NSGAII). Finally, representative simulation examples are presented to verify the proposed optimization algorithm and demonstrate its superiority over the traditional algorithm. & COPY; 2023 COSPAR. Published by Elsevier B.V. All rights reserved.
What problem does this paper attempt to address?