基于数学规划启发式的多星多站集成调度方法研究

The target observation and the data downlink to ground stations are two closely connected stages in the observation data collection process of earth observation satellites. A shortcoming of the current literature on earth observation satellite scheduling is that observation scheduling and downlink scheduling are disconnected, and the latter is usually performed after the former, therefore it is hard to achieve the holistical optimization of satellites and ground stations. In order to fully exploit the performance of the earth observation system, it is necessary to consider observation and downlink as a whole, and to manage and control satellites and ground stations at the same time. The integrated schduling of satellites and ground stations becomes inevitably a ever-crucial management science problem. This program will integrate the merits of the mathematical programming techniques, i.e., the ability of problem decomposition based on problem model, the ability of solution space exploration and the ability of solution evaluation, with the advantage of heuristics, i.e., the comparatively small computational cost in real applications. This program will carry on the research on the multi-satellite multi-station integrated scheduling method based upon matheuristics, and apply the method to the scenario of the Chinese high-resolution earth observation system. The research in this program will be of great theoretical meaning for the model and solution methodology of the complex combinatorial optimization problem arising in the field of space mission planning, moreover, the research will also be of great applicable meaning for the management and control of the high-resolution earth observation system.

对目标的观测活动和对地面站的数传活动是对地观测卫星获取观测数据的两个紧密衔接的环节。现有针对观测卫星调度研究的不足是观测过程和数传过程的调度脱节，往往在观测调度之后才进行数传调度，难以实现星地资源的整体优化。为了充分发挥对地观测系统的整体效益，必须将观测和数传作为完整的过程进行考虑，统筹兼顾卫星及地面站资源。星地资源集成调度不可避免地成为卫星对地观测应用中一个日益突出的管理科学问题。本课题结合数学规划技术基于模型的问题分解能力、解空间挖掘能力和解评价能力，考虑实际应用中启发式算法计算成本相对较低的特点，开展基于数学规划混合启发式的多星多站集成调度方法研究和以我国正在建设的高分辨率对地观测系统为背景的应用研究。本课题研究不仅在航天任务规划领域的复杂组合优化问题建模和求解算法方面具有重要的理论意义，而且对高分辨率对地观测系统的管控具有重要的实际意义。

对目标的观测活动和对地面站的数传活动是对地观测卫星获取观测数据的两个紧密衔接的环节。现有针对观测卫星调度研究的不足是观测过程和数传过程的调度脱节，往往在观测调度之后才进行数传调度，难以实现星地资源的整体优化。为了充分发挥对地观测系统的整体效益，必须将观测和数传作为完整的过程进行考虑，统筹兼顾卫星及地面站资源。星地资源集成调度不可避免地成为卫星对地观测应用中一个日益突出的管理科学问题。本课题结合数学规划技术基于模型的问题分解能力、解空间挖掘能力和解评价能力，考虑实际应用中启发式算法计算成本相对较低的特点，开展基于数学规划混合启发式的多星多站集成调度方法研究和以我国正在建设的高分辨率对地观测系统为背景的应用研究。本课题研究不仅在航天任务规划领域的复杂组合优化问题建模和求解算法方面具有重要的理论意义，而且对高分辨率对地观测系统的管控具有重要的实际意义。