基于快速筛选策略的多航天器交会任务规划  

Multiple Spacecraft Rendezvous Mission Planning Based on Fast Screening Strategy

作　　者：李明明[1] 闫慧达 黄盘兴 郭延宁[2] LI Ming-ming;YAN Hui-da;HUANG Pan-xing;GUO Yan-ning(Beijing Institute of Control Engineering,Beijing 100190 China;Harbin Institute of Technology,Harbin 150001 China)

机构地区：[1]北京控制工程研究所,北京100190 [2]哈尔滨工业大学,黑龙江哈尔滨150001

出　　处：《自动化技术与应用》2024年第8期7-11,共5页Techniques of Automation and Applications

摘　　要：研究了位于地球同步轨道(GEO)上的多航天器交会任务规划问题。以交会所需的燃料作为优化指标,建立了一个交会任务规划模型,旨在确定服务航天器的交会顺序和机动轨迹。采用灵活的两脉冲兰伯特机动策略来实现目标交会,并通过遗传算法解决优化问题。在约束条件处理中,引入了一种基于机器学习技术的分类筛选策略。快速判断个体的适应度值,减少适应度值计算,以提高进化算法的优化求解效率。最后,通过提供的仿真案例和对比实验,验证了所研究算法的卓越性能。This study addresses the issue of multiple spacecraft rendezvous mission planning in geosynchronous orbit(GEO).A rendezvous mission planning model is established with fuel consumption as the optimization criterion,aiming to determine the rendezvous sequence of the service spacecraft and the maneuver trajectory.A flexible two-impulse Lambert maneuver strategy is employed to achieve the target rendezvous,and a genetic algorithm is applied to solve the optimization problem.A classification screening strategy based on machine learning technology is introduced in dealing with constraint conditions.This strategy quickly judges the fitness value of individuals,reducing fitness value calculations,and improving the optimization efficiency of evolutionary algorithms.Finally,the excellent performance of the proposed algorithm is verified through the provided simulation cases and comparative experiments.

关 键 词：多航天器交会 兰伯特机动 分类筛选 机器学习 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] V448.2[自动化与计算机技术—控制科学与工程]