Many-objective optimization and decision-making for overall allocation of space station on-orbit activities. (December 2020)

• Record Type: Journal Article
• Authors: Zhang, Jia-cheng; Zhu, Yue-he; Luo, Ya-zhong
• Is Part Of: Acta astronautica. Volume 177(2020)
• Page Start: 202
• Page End: 216
• Publication Date: 2020-12
• Subjects: Space station -- On-orbit activity allocation -- Many-objective optimization -- Evolutionary algorithm -- Physical programming; Astronautics -- Periodicals; Outer space -- Exploration -- Periodicals; Astronautics; Periodicals; 629.405
• DOI: 10.1016/j.actaastro.2020.07.021
• Languages: English
• ISSNs: 0094-5765

Abstract

Abstract: Overall allocation of space station on-orbit activities is the process of allocating a set of on-orbit activities to be executed on the space station in a long-term planning period into several short-term subperiods while balancing logistics demand, resource requirement and the execution priority of on-orbit activities. The basic mathematical model is firstly formulated. Then a heuristic allocation strategy is proposed to reduce optimization difficulty and a hybrid decision-making approach combining physical programming and evolutionary computation is developed to directly search a Pareto-optimal solution, and the performance of these methods in addressing the problem is compared with that of NSGA-III. Three test case with 500, 1000 and 2000 on-orbit activities respectively are simulated to demonstrate the proposed approaches. The results show that the heuristic allocation strategy is effective and performs better than only applying meta-heuristics, and the decision-making approach based on physical programming can efficiently achieve a decision-maker-preferred solution that is converged better than the Pareto frontier obtained by NSGA-III. The obtained allocation scheme satisfies all equilibrium indexes and constraints. Highlights: The mathematical model of overall allocation of space station on-orbit activities is constructed. The heuristic allocation strategy is proposed to reduce the number of objective functions. A hybrid decision-making approach is developed to obtain the desirable solution efficiently.