基于非关系型数据库的卫星参数存储与处理

Storage and Processing of Satellite Engineer Data Based on NoSQL Database

随着空间科学探测计划以及各个空间探测先导专项的实施,越来越多的空间科学卫星得以立项和升空,包括暗物质粒子卫星,以及实践十号卫星,还有量子科学实验卫星等等。传统的关系型数据已不太适合针对卫星数据的纵向比较和处理,甚至是对其有效载荷的工作情况进行评估和预测。对此,提出了一种以非关系型数据库为主导的卫星工程参数存储与处理的研究方式,使用当前比较流行的三种不同类型的四种非关系型数据库存储卫星工程数据,以及使用多项式拟合外推,求和自回归滑动平均模型,非参数回归算法以及基于神经网络的预测技术等四种数据处理预测方法来针对当前工程数据进行分析处理。给出了一种比较好的处理方案,以迎合越来越蓬勃的空间科学卫星的发展计划,为空间科学卫星的地面支撑系统建设提供一种思路。

With development of space science exploration and advance leading projects,more and more space science satellites are designedand launched,including dark matter particle explorer （DAMPE）,SJ-10,and quantum experiments at space scale （QUESS）. The traditionarelational database is not suitable for the data comparison and processing,besides the assessing and prediction on working states of payload.For this,we present a method for the storage and processing of satellite engineering parameters dominated by non-relational database and usfour common non-relational database out of three different types for storage of satellite engineering data and four data processing forecastingmethods including polynomial fitting,ARMA,non -parametric regression and back propagation network to analyze the engineering data.There we give a better scheme to adapt to the development plan of more vigorous space science satellites,which provides a way for the construction of the ground support system of space science satellites.