基于BFA的SPOT5卫星日常拍摄任务调度算法  

BFA-based algorithm for the SPOT5 daily photograph scheduling problem

作　　者：刘祥 罗开平[1] 王海红 郭清峰[1] LIU Xiang;LUO Kaiping;WANG Haihong;GUO Qingfeng(School of Economics and Management,Beihang University,Beijing 100191,China;Beijing Institute of Tracking&Telecommunication Technology,Beijing 100094,China)

机构地区：[1]北京航空航天大学经济与管理学院,北京100191 [2]北京跟踪与通信技术研究所,北京100094

出　　处：《系统工程理论与实践》2020年第3期783-794,共12页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(71871004,91646110)。

摘　　要：地球影像需求日益旺盛,如何最优化调度日常成像卫星成为了一个亟待解决的问题.SPOT5是一颗典型的商用成像卫星,针对其日常成像调度问题的特点,开发了一个带修补策略的二进制萤火虫算法.该算法通过修补策略的丢项操作修复不可行解,通过修补策略的增项操作提升解的质量.修补策略优先处理具有高度稀疏性的逻辑约束,再处理信息存储容量约束,从而增强算法在大规模组合空间的寻优能力.在公开的SPOT5卫星基准测试问题集上的实验结果验证了所提算法的高效性,而且所得解的质量显著好于遗传算法的运行结果.The scheduling of daily photograph satellites has obtained more and more attentions as the requirement for earth image steadily increases.SPOT5 investigated in this paper is a typical commercial imaging satellite.A binary firefly algorithm with a repairment strategy is developed in term of the feature of the SPOT5 daily photograph scheduling problem.In the repairment strategy of the proposed algorithm,the dropping operation repairs the infeasible solution and the adding operation improves the quality of solution.The proposed repairment strategy firstly handles the logical constraints with a large number of sparse coefficients and then the onboard memory limitation to enhance the searching ability of the proposed algorithm in the large-scale combinatory space.The experimental results on the SPOT5 benchmark test suite verify the efficiency of the proposed algorithm and its superiority to the genetic algorithm in term of the quality of the solution found.

关 键 词：优化 卫星调度 二进制萤火虫算法 SPOT5 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]