超额订购下天文卫星短期任务的聚类规划方法  

Clustering-Scheduling Methods for Oversubscribed Short-Term Tasks of Astronomical Satellites

作　　者：尹晓丹 白萌 李卓恒 YIN Xiaodan;BAI Meng;LI Zhuoheng(Laboratory of Satellite Operations Technology,Space Science Mission Operations Center,National Space Science Center,Chinese Academy of Sciences,Beijing 100190,P.R.China;University of Chinese Academy of Sciences,Beijing 100049,P.R.China)

机构地区：[1]中国科学院国家空间科学中心卫星测运控技术实验室,北京100190 [2]中国科学院大学,北京100049

出　　处：《Transactions of Nanjing University of Aeronautics and Astronautics》2023年第3期307-322,共16页南京航空航天大学学报（英文版）

基　　金：supported by the Strategic Priority Research Program of the Chinese Academy of Sciences(No.XDA15040100);the Youth Innovation Promotion Association of the Chinese Academy of Sciences(No.2021146).

摘　　要：当用户的观测需求超过卫星的观测能力时,天文卫星的任务规划就成为一个超额订购的问题。对于该问题,设计了一个包含聚类阶段和短期任务规划阶段的框架。首先建立了任务聚类模型,用于减少超额订购任务的规模。其次,使用聚类的结果作为输入,建立了短期任务规划的数学模型。最后,提出了一种自适应混合搜索策略的人工蜂群算法,在基本人工蜂群算法中引入了自适应精英全局⁃局部搜索策略和自适应变邻域最优搜索策略,以求解聚类和短期规划问题。所提出的算法在实验中表现出更好的寻优能力和更快的收敛速度。此外,它还有效地减少了聚类阶段的任务数量,提高了短期任务规划阶段的任务完成度。When the observation requirement from users exceeds the satellite’s observation capability,astronomy satellite task scheduling becomes an oversubscription problem.For the oversubscribed task scheduling of astronomical satellites,a framework with a clustering phase and a short-term task scheduling phase is designed.First,a task clustering model is established to reduce the size of the oversubscribed task.Second,using the clustered results as input,we develop a mathematical model of short-term scheduling for the tasks.Finally,we propose an improved artificial bee colony algorithm with adaptive hybrid search strategies(DirectABC).It introduces an adaptive elite global-local search strategy and an adaptive variable neighborhood optimal search strategy to the basic artificial bee colony algorithm(BasicABC).The proposed algorithm demonstrates superior optimum-searching capability and a faster convergence speed in the simulation.In addition,it effectively reduces the number of tasks in the clustering phase and improves task completion in the short-term task scheduling phase.

关 键 词：天文卫星任务规划 超订购问题 任务聚类 短期任务规划 人工蜂群算法 

分 类 号：V474[航空宇航科学与技术—飞行器设计]