To solve the multi-objective optimization problem for attitude maneuver of liquid-filled flexible spacecraft, the authors propose a novel algorithm based on an improved hierarchical structure. This improved hierarchical optimization algorithm (IHOA) is composed of elitist non-dominated sorting genetic algorithm (NSGA-II) and bare-bones multi-objective particle swarm optimization based on r-dominance relation (r-BBMOPSO). Bottom layer algorithm NSGA-II provides elitist individuals for top layer algorithm. Top layer algorithm r-BBMOPSO employs r-domination instead of Pareto domination to strengthen selection pressure and guide the search toward the decision maker’s preference. The modification of hierarchical structure has reduced the influence of randomness and accelerated convergence speed while preserving the operating efficiency of the algorithm. The feasibility and effectiveness of IHOA are demonstrated by bench-mark test problems. Also, the simulation results proved that the proposed optimization algorithm IHOA has the capacity to stably generate solutions which are satisfied the decision maker’s demand in the case of a single run. Especially, based on the optimized parameters, the spacecraft can perform well in attitude maneuver tasks.

为了解决充液柔性航天器姿态机动的多目标优化问题，作者提出了一种基于改进层次结构的新算法。该改进的层次优化算法（IHOA）由精英非支配排序遗传算法（NSGA-II）和基于r主导关系的裸骨多目标粒子群算法（r-BBMOPSO）组成。底层算法NSGA-II为顶层算法提供了精英人士。顶层算法r-BBMOPSO采用r控制而不是帕累托控制，以增强选择压力并引导搜索朝决策者的偏爱方向发展。分层结构的修改减少了随机性的影响，加快了收敛速度，同时保留了算法的运行效率。基准测试问题证明了IHOA的可行性和有效性。仿真结果还表明，所提出的优化算法IHOA具有稳定生成能够满足决策者单次运行需求的解决方案的能力。特别地，基于优化的参数，航天器可以在姿态机动任务中很好地执行。