一种流程工业的分布式海量报警管理系统  

A Distributed Mass Alarm Management System in Process Industry

作　　者：胡静泓 李德文[1] 黄文君[1] 杨奇泽 

机构地区：[1]浙江大学控制科学与工程学系,杭州310027 [2]浙江中控技术股份有限公司,杭州310053

出　　处：《上海交通大学学报》2015年第11期1660-1664,1669,共6页Journal of Shanghai Jiaotong University

基　　金：国家自然科学基金(61273100)资助项目

摘　　要：提出了一种应用于流程工业的分布式海量报警管理系统,利用Hadoop平台的分布式可扩展特性实现了海量报警数据的可靠存储和高效处理.在数据存储层,通过合理的列式存储结构设计,实现了海量报警数据的分布式集群存储;在信息查询层,实现了类MapReduce过程的报警管理系统分布式并行查询模型;在功能应用层,根据业务需求实现了基本的报警管理功能模块.测试结果表明,面对海量报警数据的应用场景,基于Hadoop平台的分布式海量报警管理系统比传统报警管理系统具有更为出色的性能.Adistributed mass alarm management system was proposed in process industry.The reliable storage of mass alarm and efficient computing was achieved by utilizing extensible and distributed features of Hadoop.Reasonable design of column type storage structure made it possible to realize,distributed clustered storage of mass alarm data in data storage layer.In information processing layer,distributed parallel query model of alarm management system like MapReduce was achieved.In functional layer,according to the requirement of the system,basic alarm management function modules were implemented.The test results show that the distributed mass alarm management system based on Hadoop works better than the traditional alarm management system when facing application scenario with mass alarm data.

关 键 词：海量报警管理 流程工业 HADOOP 列式存储结构设计 分布式集群存储 分布式并行查询模型 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]