Reactive scheduling of multiple EOSs under cloud uncertainties: model and algorithms

• 作者: WANG Jianjiang HU Xuejun HE Chuan
• 来源期刊: 系统工程与电子技术（英文版）
• 年，卷（期）: 2021,（1）
• 所属期刊栏目: SYSTEMS ENGINEERING
• 页码范围: 163-177
• 页数: 15页
• 语种: 英文
• DOI: 10.23919/JSEE.2021.000015

摘要

Most earth observation satellites (EOSs) are low-orbit satellites equipped with optical sensors that cannot see through clouds. Hence, cloud coverage, high dynamics, and cloud un-certainties are important issues in the scheduling of EOSs. The proactive-reactive scheduling framework has been proven to be effective and efficient for the uncertain scheduling problem and has been extensively employed. Numerous studies have been conducted on methods for the proactive scheduling of EOSs, in-cluding expectation, chance-constrained, and robust optimiza-tion models and the relevant solution algorithms. This study fo-cuses on the reactive scheduling of EOSs under cloud uncertain-ties. First, using an example, we describe the reactive schedul-ing problem in detail, clarifying its significance and key issues. Considering the two key objectives of observation profits and scheduling stability, we construct a multi-objective optimization mathematical model. Then, we obtain the possible disruptions of EOS scheduling during execution under cloud uncertainties, adopting an event-driven policy for the reactive scheduling. For the different disruptions, different reactive scheduling algorithms are designed. Finally, numerous simulation experiments are con-ducted to verify the feasibility and effectiveness of the proposed reactive scheduling algorithms. The experimental results show that the reactive scheduling algorithms can both improve obser-vation profits and reduce system perturbations.