支持大规模监测数据分析的LaUD-KV聚集操作扩展  

LaUD-KV aggregation function expansion for massive equipment monitoring data analysis

作　　者：徐昊[1] 杨义繁[1] 张力[1] 刘英博[1] 王建民[1] 

机构地区：[1]清华大学软件学院,北京100084

出　　处：《计算机集成制造系统》2013年第12期3035-3042,共8页Computer Integrated Manufacturing Systems

基　　金：国家863计划资助项目(2012AA040911);国家科技支撑计划资助项目(2012BAF12B23)~~

摘　　要：结合当前维护、维修和大修行业对于大数据存储和查询的需求,对当前工业数据存储方案和大数据查询模式的优缺点进行了研究和分析,提出一套在分布式存储环境下工业大数据的聚集函数查询框架和对应算法。基于现有的LaUD-KV存储模式和LaSQL语言,实现了一套能满足工业大数据查询需求的类SQL聚集函数,使得在数据过亿的情况下,系统对于聚集查询的响应时间缩短为原有的30%左右,同时为上层的维护、维修和大修监测平台提供了技术支持。For the requirement of Maintenance, Repair and Overhaul/ Operation （MRO） industry for storage and query of mass data, the advantages and disadvantages of existed storage scheme and query mode were studied. A new framework for aggregation functions and its corresponding algorithms base on distributed storage environment was proposed. On the basis of existing LaUD-KV storage engine and LaSQL query language, a set of SQL aggrega tion function which satisfied the industrial data query was realized. With more than billion data, the response time of system for aggregation query was reduced by 70%, and the technical support for upper MRO monitoring system was provided.

关 键 词：维护 维修和大修 监测数据 LaUD-KV 聚集函数 查询框架 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]