一种航天器交会与接近路径规划算法  

A Path Planning Algorithm for Spacecraft Rendezvous and Approach

作　　者：黄成 王涛 许家忠[1] HUANG Cheng;WANG Tao;XU Jiazhong(School of Automation,Harbin University of Science and Technology,Harbin 150080,China)

机构地区：[1]哈尔滨理工大学自动化学院,哈尔滨150080

出　　处：《宇航学报》2023年第8期1225-1237,共13页Journal of Astronautics

基　　金：国家自然科学基金(52102455)。

摘　　要：针对非合作交会与接近任务中近程导引段航天器的三维路径规划问题,提高其在多约束条件下的寻优性能,提出一种融合混沌映射、自适应随机动态扰动、随机探索机制以及反向学习策略的混合蝙蝠算法。首先,利用Circle混沌映射进行种群初始化,提高种群多样性;在全局搜索中引入自适应随机动态扰动系数,并在位置更新时进行位置限定,提高算法全局寻优的性能;在局部搜索中融入柯西分布的逆累积分布函数,避免算法进入局部最优。同时,引入分段随机的反向学习策略,进一步提高种群多样性,平衡全局寻优与局部探索的性能。然后,以基于升序排列的路径节点进行三次样条插值编码,构造躲避障碍物求解航天器三维路径的适应度函数,结合混合蝙蝠算法求得平滑路径曲线。最后,通过对比仿真表明了所提方法的最优路径次数较对比算法增加,最差路径次数减少,路径平均解减少。Aiming at the three-dimensional path planning problem of the spacecraft in the short range guidance phase of the non-cooperative rendezvous and approach mission,to improve its optimization performance under multiple constraints,a hybrid bat algorithm integrating chaotic mapping,adaptive random dynamic disturbance,random exploration mechanism and reverse learning strategy is proposed.Firstly,Circle chaotic mapping is used for population initialization to improve population diversity.The adaptive stochastic dynamic perturbation coefficient is introduced into the global search,and the position is limited when the position is updated to improve the performance of the global search algorithm.The inverse cumulative distribution function of Cauchy distribution is incorporated into local search to avoid local optimization.At the same time,the segmented random reverse learning strategy is introduced to further improve the population diversity and balance the performance of global optimization and local exploration.Then,the ascending path nodes are coded with cubic spline interpolation to construct the fitness function of obstacle avoidance to solve the three-dimensional path of the spacecraft,and the smooth path curve is obtained by combining the hybrid bat algorithm.Finally,comparison simulation shows that the optimal path time of the proposed method is increased,the worst path time is reduced and the average path solutions are reduced.

关 键 词：航天器 非合作交会 蝙蝠算法 混沌映射 反向学习 自适应随机探索 

分 类 号：V249[航空宇航科学与技术—飞行器设计]