航天控制系统中国产化海量数据处理优化技术  

The Optimization Technique of Mass Data Processing Based on the Domestic Platform of Aerospace Control System

作　　者：刘孟语 张洁[1] 上官子粮 赵玉梅[1] 杨喆[1] Liu Mengyu Zhang Jie Shangguan Ziliang Zhao Yumei Yang Zhe(Beijing Aerospace Automatic Control Institute, Beijing 100854, China)

机构地区：[1]北京航天自动控制研究所,北京100854

出　　处：《航天控制》2016年第6期57-63,共7页Aerospace Control

摘　　要：信息化的发展使得航天武器装备在控制系统试验和使用中产生了海量数据。针对国外设备安全性无法保障和国产化软硬件平台性能不足的问题,本文提出一种基于国产化平台的海量数据处理优化技术,确保数据处理系统可以安全、快速地处理控制系统的试验数据。首先,采用多处理技术和分布式并行计算2种方式,提升系统数据计算能力。随后,采用最小单元UNIT模式对数据进行分级抽样等预处理,减小数据入库、出库的压力。数据查询实验结果表明,系统的数据查询速率有了大幅提升,且实现了任意试验时间段数据的秒级加载。With the development of information, Mass data are produced while the aerospace control system is tested and used. According to the insecurity of foreign equipment and the low performance of domestic server, a data processing and optimization technique based on domestic server is implemented, which ensures the safety and fleetness of domestic DPS （Data Processing System） for control system. Firstly, multi-processing, distributed and parallel computing two techniques are implemented to upgrade the data processing capacity. Then, in order to decrease the pressure of data-input and data-output, the minimum UNIT mode is applied to preprocess data. The result of data query experiments shows that the speed of data query is increased greatly and the data in arbitrary test period can be loaded in second level.

关 键 词：海量数据 国产化数据处理系统 多处理技术 分布式并行处理 最小单元 分级抽样 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]