航空学报 > 2024, Vol. 45 Issue (16): 329706-329706   doi: 10.7527/S1000-6893.2023.29706

基于滚动时域策略的中继卫星多目标动态调度优化方法

李恒伟1, 罗启章1, 顾轶1, 范才智2, 伍国华1()   

  1. 1.中南大学 交通运输工程学院,长沙 410075
    2.国防科技大学 空天学院,长沙 410073
  • 收稿日期:2023-10-12 修回日期:2023-10-28 接受日期:2023-12-22 出版日期:2024-01-15 发布日期:2024-01-11
  • 通讯作者: 伍国华 E-mail:guohuawu@csu.edu.cn
  • 基金资助:
    国家自然科学基金(62073341)

Multi⁃objective dynamic scheduling optimization method for relay satellites based on rolling horizon strategy

Hengwei LI1, Qizhang LUO1, Yi GU1, Caizhi FAN2, Guohua WU1()   

  1. 1.College of Traffic and Transportation Engineering,Central South University,Changsha 410075,China
    2.College of Aerospace Science and Engineering,National University of Defense Technology,Changsha 410073,China
  • Received:2023-10-12 Revised:2023-10-28 Accepted:2023-12-22 Online:2024-01-15 Published:2024-01-11
  • Contact: Guohua WU E-mail:guohuawu@csu.edu.cn
  • Supported by:
    National Natural Science Foundation of China(62073341)

摘要:

为提高中继卫星系统应对突发状况的能力,提出了基于滚动时域策略的中继卫星动态调度方法,将复杂动态调度过程分解为数个静态调度子问题,同时根据中继卫星动态调度需求,设计了多目标优化算法对子问题进行求解。首先构建了以最大任务完成率和最小调度方案调整幅度为优化目标的中继卫星动态调度模型,然后基于动态调度特性提出了动态任务调度方法,该方法采用周期和事件驱动的混合重调度机制,将调度过程分成多个调度区间,在每个区间内利用自适应邻域搜索的多目标进化算法对窗口任务进行调度。为了验证所提出的动态调度模式及算法的有效性,开展了大量仿真实验。实验结果证明了所提方法在解决中继卫星动态调度问题方面的优越性,与前沿的多目标优化方法NSGA-Ⅱ、MDSA-NSGA-Ⅱ和MODJA相比,本文设计的算法能生成更高质量的解。

关键词: 中继卫星, 动态调度, 滚动时域策略, 多目标优化, 自适应邻域结构

Abstract:

To improve the ability of responding to emergencies in the scheduling of relay satellite system, a dynamic scheduling model is proposed based on the rolling horizon strategy in this paper, which decomposes the complex dynamic scheduling process into several static scheduling subproblems. A multi-objective optimization algorithm is designed to solve the subproblems according to the demand for dynamic scheduling of the relay satellite. Firstly, a dynamic scheduling model for the relay satellite is constructed to obtain the maximum task completion rate and the minimum adjustment range of scheduling scheme. Then, based on dynamic scheduling characteristics, a dynamic task scheduling method is proposed. The method adopts a hybrid rescheduling mechanism based on cycle and event-driven, divides the scheduling process into scheduling intervals, and uses a multi-objective evolutionary algorithm based on adaptive neighborhood search to schedule the window tasks in each interval. To verify the effectiveness of the proposed dynamic scheduling model and algorithm, a large number of simulation experiments are carried out. The experimental results prove the superiority of the proposed method in solving the dynamic scheduling problem of relay satellite. Compared with the cutting-edge multi-objective optimization methods of NSGA-Ⅱ, MDSA-NSGA-Ⅱ, and MODJA, the algorithm designed in this paper can generate higher quality solutions.

Key words: relay satellite, dynamic scheduling, rolling horizon strategy, multi-objective optimization, adaptive neighborhood structure

中图分类号: