【作者】 白杰；

【导师】 马萍；

【作者基本信息】 哈尔滨工业大学 ， 控制科学与工程， 2025， 硕士

【摘要】 低轨卫星星座凭借轨道高度低、传输时延小、覆盖范围广等优势,在全球通信、遥感监测等领域具有重要应用价值。随着卫星组网规模不断扩大,星座系统可靠性面临严峻挑战——低轨卫星因空间环境复杂、寿命周期短等特点失效率显著高于传统中高轨卫星。为确保星座持续稳定运行,设计科学有效的维护方案至关重要。备份策略作为目前主要的维护手段,通常在地面或空间备份节点部署备用卫星以实现对故障卫星的快速替换,但单一备份方式难以满足大规模星座的实际需求。为此,构建多级融合的备份策略成为解决低轨大规模星座维护难题的关键。本文以低轨大规模星座多级备份策略为核心,对地面卫星发射任务规划、备份卫星轨道转移和多级备份资源配置三个重要环节进行研究。首先,针对地面备份卫星补充发射的多中心多任务协同规划问题,提出不同模式下建模与求解方法。面向单星发射模式,构建以最小化未完成加权任务数为目标的混合整数规划模型,采用遗传算法与序列二次规划分层求解任务分配与发射时序;针对多星发射场景,建立同时考虑任务完成率与发射成本的多目标优化模型,设计基于贪婪原则的大邻域搜索算法,通过破坏-修复机制迭代优化方案,并引入精英集合维护机制保存Pareto前沿解集。通过仿真实验分析不同任务规模下发射模式的选择对发射系统效能的影响。其次,针对备份卫星从停泊轨道至目标轨道的转移优化问题,建立以燃料消耗最小化为目标的多脉冲转移模型,提出改进鲸鱼算法(Mod-WOA)。算法改进包括两方面:采用混沌映射与反向学习策略增强初始种群多样性,避免早熟收敛;引入基于多臂赌博机理论的自适应选择算子,平衡算法全局探索与局部搜索能力。构建基准测试问题集和典型轨道转移场景验证算法改进效果以及多脉冲转移策略的实用价值。最后,提出融合地面快速补发、停泊轨道预部署及在轨热备份的三级备份方案,构建以年备份成本最小化为目标的资源配置模型,综合考虑火箭运载能力、发射中心能力限制等约束。基于(s,Q)库存管理理论建立系统效率评估框架,采用边际分析法定量分析关键参数对备份成本及系统效率的影响。以实际大规模星座为案例进行仿真,在统一效率指标要求下,对比三级备份方案和传统二级备份方案的资源配置方案,体现三级备份策略的优势。更多还原

【Abstract】 Low-earth orbit(LEO)satellite constellations,by virtue of their low orbital altitude,short transmission delay,wide coverage and other advantages,have significant application values in such fields as global communications and remote sensing and monitoring.As the scale of satellite networking continues to expand,the reliability of the constellation system faces serious challenges,and since LEO satellites are characterized by a complex space environment and a short life cycle,their failure rate is significantly higher than that of traditional medium-and high-orbit satellites.In order to ensure the sustainable and stable operation of constellations,it is crucial to design a scientific and effective maintenance program.As the main means of maintenance at present,the backup strategy usually adopts the three modes of ground launch,same-orbit backup and parked orbit replenishment,all of which belong to the single backup mode,however,it is difficult for the single backup mode to satisfy the actual needs of large-scale constellations.Therefore,the construction of a multi-level backup strategy system has become the key to solve the maintenance problems of LEO large-scale constellations.In this paper,taking the optimization of multilevel backup strategy for LEO large-scale constellations as the core,three important links,including launch mission planning,orbit transfer and backup resource allocation,are systematically investigated.First,for the multi-center and multi-task cooperative planning problem in the supplementary launch of ground backup satellites,modeling and corresponding solution methods are proposed under different modes.For the single-satellite launch mode,a mixed-integer planning model is constructed with the objective of minimizing the weighted number of uncompleted tasks,and a hierarchical solution framework combining genetic algorithms and sequential quadrat ic programming is adopted for step-by-step solution with respect to the task assignment and launch timing;for the multi-satellite launch mode,a multi-objective model considering both the task completion rate and the launch cost is established,and a large-neighborhood search algorithm based on the greedy principle is designed for establishing an optimization scheme with iterative destruction-repair mechanism,and introduce the elite set maintenance mechanism to preserve the Pareto frontier solution set.The high efficiency of the single-star model and the economy of the multi-star model are verified through simulation experiments of different scales,respectivelySecond,for the transfer optimization problem of the backup satellite from the parking orbit to the target orbit,a multi-impulse transfer model with the goal of minimizing fuel consumption is established,and the Modified Whale Optimization Algorithm(Mod-WOA)is proposed.The algorithm improvement includes two aspects:on the one hand,the improved algorithm adopts chaotic mapping and inverse learning strategy to enhance the diversity of the initial population and avoid premature convergence;on the other hand,the improved algorithm introduces an adaptive parameter adjustment mechanism based on the theory of multi-armed gaming machine,which balances the algorithm’s global search and local optimization ability.Through the validation in typical orbital rendezvous scenarios,the multi-pulse transfer strategy significantly reduces the fuel consumpti on compared with the traditional double-pulse approach,which reflects the engineering practicability of the improved algorithm.Finally,a three-level backup strategy is proposed,which includes the integration of ground rapid replenishment,parking orbit pre-deployment and in-orbit hot backup,and an optimization model is constructed with the goal of minimizing the annual backup cost,taking into account the constraints of rocket capacity and launch center capacity limitations.Based on the inventory mana gement theory,a system efficiency evaluation framework is established,and the influence of key parameters on backup cost and system efficiency is quantitatively analyzed by the marginal analysis method,and simulation is carried out using an actual large-scale constellation as a case study.The results show that the three-stage backup strategy can effectively reduce the total cost of the backup system under the premise of meeting the efficiency requirements,which verifies the advantages of the multi-level backup strategy.更多还原