基于多目标遗传算法的卫星成像任务调度技术  

Scheduling techniques of satellite imaging tasks based on multi-objective genetic algorithm

作　　者：李军[1] 王钧[1] 陈健[2] 陈慧中[1] 郭玉华[1] 

机构地区：[1]国防科技大学电子科学与工程学院,湖南长沙410073 [2]北京遥感信息研究所,北京100085

出　　处：《系统工程与电子技术》2007年第7期1164-1168,共5页Systems Engineering and Electronics

基　　金：国家自然科学基金资助课题(60604035)

摘　　要：针对成像卫星的任务调度问题,考虑多种约束条件,提出了成像任务约束图模型。基于多目标优化和遗传算法思想,提出了一种应用于卫星成像任务调度的多目标遗传算法,详细分析了算法的各个关键步骤的设计思想,能够求得基于成像约束图模型的满意解,生成卫星的优化拍摄计划。卫星成像任务调度原型系统的实现和实验结果验证了该模型和算法的合理性和有效性。A new imaging task constraint graph model is proposed for planning and scheduling of satellite imaging tasks with many constraints. Based on multi-objective optimization and genetic algorithm, a new multiobjective genetic algorithm is proposed for planning and scheduling of satellite imaging tasks, and the key steps of the algorithm are presented. The proposed algorithm can be computed for satisfaction solution of imaging task constraint graph, thus optimized imaging plans of are produced. On this basis, the prototype system for planning and scheduling of satellite imaging tasks is implemented, and many experiments are completed. Accordingly its rationality and validity are proved.

关 键 词：成像任务调度 多目标遗传算法 成像约束图 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]