摘要
卫星任务规划问题的求解空间随卫星数量和目标数量的增多而快速增大,针对大规模多星多任务规划问题,提出一种基于关键路径-遗传算法的卫星任务规划分层优化方法。该方法将卫星任务规划问题分解成任务分配和单星任务处理2个子问题,其中,任务分配通过遗传算法实现,一个分配结果对应种群中的一个个体,在单星任务处理子问题中引入关键路径法,用于求解每个个体的适应度,在提高优化效率的同时,确保得到当前任务分配条件下的最大观测效益,提高算法的全局优化能力。仿真结果表明,对于给定的6组大规模卫星任务规划算例,平均任务完成率均超过了99.7%,证明了该方法具有良好的稳定性和全局搜索能力;同时,相比于已有方法,该方法在优化效率上也有大幅提高,且任务规模越大,优化效率提升越大。
With the increase of the number of satellites and targets, the space for solving satellite task planning increases rapidly. For large-scale multi-satellite and multi-task planning, a hierarchical optimization method, which consists of Critical Path Method(CPM) and Genetic Algorithm(GA) is proposed. The method decomposes the satellite task planning into two subproblems: task allocation and single-satellite task processing. Task allocation is realized by GA and a task allocation result corresponds to an individual of the population. The CPM is used in the single-satellite task processing to solve the fitness of each individual, which improves the optimization efficiency, ensures the maximum observation benefit under the current task assignment condition, and improves the global optimization ability of the algorithm. Simulation results show that for a given set of 6 examples, the mission completion rate is over 99.7%, which proves that the method has good stability and global searching ability. Compared with the existing methods, the method also enhances the optimization efficiency greatly, and the larger the mission scale is, the higher the optimization efficiency will be improved.
作者
毛李恒
邓清
刘柔妮
孔祥龙
Mao Liheng;Deng Qing;Liu Rouni;Kong Xianglong(Shanghai Institute of Satellite Engineering,Shanghai 201109,China)
出处
《系统仿真学报》
CAS
CSCD
北大核心
2021年第1期205-214,共10页
Journal of System Simulation
基金
国防基础科研项目(JCKY2016203A017)。
关键词
卫星调度
任务规划
分层优化
关键路径法
遗传算法
satellites scheduling
task planning
hierarchic optimization
critical path method
genetic algorithm
