量子科学实验卫星的多维数据立方体方法研究

Research on the Multidimensional Data Cube Method of the Quantum Science Experiment Satellite

科学实验卫星以科学目标为任务战略导向,需要任务团队结合实际业务运行情况制定阶段实验计划.这些实验计划以任务运行数据为依据,由科学卫星任务各分系统数据分析获得的策略集合综合得到.量子科学实验卫星在轨运行期间会产生大量运行数据,如何有效利用这些数据给不同决策层提供辅助是当前面临的一大难题.目前主流方式主要借助值班日志统计和常规的数据库系统数据统计,需耗费较多时间和人力成本,对分析人员专业技能要求较高,无法满足多角度和多粒度任务研判的要求,并且该方法可扩展性差,当问题的观察角度变化时,往往需要重新组织数据统计分析.针对上述问题,本文提出一种基于数据立方体的多维数据建模和分析方法,对于不同主题,能够对数据进行多层次、多角度、多粒度统计分析,为决策提供良好支持.

Scientific experiment satellite should be target-oriented,which requires several mission teams to develop a phase experiment plan and take operations according to the actual situation.These experiment plans depend on the data generated by mission operation and are integrated from the strategy set derived from some data analysis of each subsystem of the scientific satellite mission.Along with the operation of QUESS on orbit,a large amount of operational data will be generated.To solve this problem,current mainstream approach mainly relies on log statistics and data statistics of conventional database systems.However,these methods consume more energy and time,and require more professional skills of analysts,which cannot meet the requirements of multi-angle and multi-granularity research and judgment tasks.Moreover,the scalability of the methods is very poor.When the observation angle of the problem changes,it is often necessary to reorganize the statistical analysis of the data.In view of the above situations,a multi-dimensional data modeling and analysis method based on data cube is proposed,which can face different topics and support multi-level,multi-angle and multi-granularity statistical analysis of data,providing good support for decision makers.