面向收益不均等区域目标的多星调度问题

利用多星对区域目标的协同观测在工程应用方面的需求日益复杂化,当目标区域内待观测目标的重要性不同,其对应的观测收益不同。那么如何在资源有限的情况下,安排观测方案使观测总收益最大是亟待解决的问题。考虑区域目标内部观测收益不均等的实际情况,建立了区域目标内部收益随时间可变的多星调度优化模型,设计了条带动态分割方法(dynamic stripe segmentation method, DSSM),提出了基于权重优先策略的启发式算法(weight priority strategy heuristic, WPSH)和基于随机邻域的局部搜索算法(random neighborhood local search, RNLS),使求解结果对应的观测收益最大,最后通过仿真实验验证DSSM算法比并行分割算法观测收益提高约19%,覆盖率提高约22%;RNLS算法比WPSH算法的基础上平均提升观测收益约4.5%,且求解质量较稳定。

Multi-satellite scheduling problem for regional targets with uneven income

The need for collaborative observation of multi-satellite regional targets is becoming increasingly complex in engineering applications. However, due to differences in geological structure and character activities in a target to be observed, the importance of the targets to be observed in the target is different, the corresponding observed returns are different. How to arrange the observation program to maximize the total benefit of the observation is a problem that needs to be solved in the case of limited resources. In this paper, considering the actual situation of unequal internal observation income of regional targets, a multi-satellite scheduling optimization model with variable internal income of regional targets is established. The dynamic stripe segmentation method (DSSM) is designed and proposed based on weights. The weight priority strategy heuristic (WPSH) and the random neighborhood local search (RNLS) algorithm maximize the observation benefit corresponding to the solution result. Finally, the simulation experiment proves that the DSSM algorithm improves the observation yield by about 19% and the coverage rate by about 22%. The RNLS algorithm is about 4.5% higher than the WPSH algorithm, and the solution quality is stable.