摘要
针对预警系统卫星资源调度问题,提出了一种多星协同的分布式策略.该策略依据扫描得到的目标弹道信息,通过调度多颗低轨卫星实现对目标的接力探测、协同定轨和持续跟踪.本研究的核心在于开发了一种完全分布式的调度方法,该方法基于协同进化遗传算法,使每颗卫星具备独立探索和优化自身调度方案的能力,彻底摆脱了对中心节点的依赖.此外,引入了一种基于接收、更新与传播的多星协商机制,确保在分布式环境下信息认知的一致性.通过仿真实验验证,该方法能够获得多星一致认可的最优调度方案,并显著减少任务切换次数.这一成果不仅提高了任务执行的效率,还极大地增强了系统的稳定性和可靠性,为卫星资源调度提供了新的有效解决方案.
Low Earth orbit(LEO)early warning systems are primarily used for tracking and identifying the mid-course phase of ballistic missiles.Early warning constellation mission scheduling refers to the process of using high Earth orbit(HEO)satellites to scan and obtain target ballistic missile information,and then scheduling multiple LEO satellites to conduct relay detection of the target over a period of time.However,with the continuous increase in the number of satellites in orbit,the entire system becomes increasingly complex.The rise of hypersonic weapons also poses a significant challenge to early warning systems.How to efficiently schedule on-board detection resources to achieve early warning tasks for targets is a complex and critical issue.This paper proposes a fully distributed method without a central node,combining co-evolutionary genetic algorithms with mission scheduling models to optimize the problem in a distributed manner,achieving autonomous on-board scheduling oriented towards early warning tasks.The multi-satellite scheduling problem is transformed into a series of single-satellite scheduling problems,where each satellite independently explores its own scheduling sub-problems.The fitness of a single-satellite plan,combined with other satellite plans to form the constellation’s overall scheduling plan,is evaluated using indicators such as satellite contribution,sensor switch count,satellite load,and early warning benefits,thereby optimizing the local scheduling plan for each satellite.To ensure consistency in information cognition,global negotiation communication is required within the constellation.A ring topology structure is used to form an inter-satellite communication network,and the multi-satellite negotiation mechanism in a distributed environment is discussed.A star-based negotiation strategy based on receive,update,and send is adopted.Each satellite continuously sends its current latest plan to neighboring satellites and optimizes and updates its own plan based on the latest plans of other satellites received from neighboring satellites.Simulation results show that this method can converge to an optimal scheduling plan recognized by multiple satellites within a certain time,effectively reducing the number of switches.
作者
李佳雯
叶东
肖岩
张梓毅
LI JiaWen;YE Dong;XIAO Yan;ZHANG ZiYi(Research Center of Satellite Technology,Harbin Institute of Technology,Harbin 150080,China;State Key Laboratory of Micro-Spacecraft Rapid Design and Intelligent Cluster,Harbin 150080,China)
出处
《中国科学:物理学、力学、天文学》
北大核心
2025年第9期90-101,共12页
Scientia Sinica Physica,Mechanica & Astronomica
基金
国家自然科学基金(编号:62073102,62203145)
国家重点研发计划(编号:2021YFC2202900)
中国博士后科学基金(编号:2022M710948)
民用航天技术预先研究(编号:D030312)资助项目。
关键词
星座预警
任务调度
多星协同
分布式优化
constellation warning
task scheduling
co-evolution
distributed optimization
作者简介
联系人:叶东,E-mail:yed@hit.edu.cn。
