【作者】 屈志坚； 郭亮； 刘明光； 陈秋琳； 蒋俊俊；

【Author】 QU Zhijian;GUO Liang;LIU Mingguang;CHEN Qiulin;JIANG Junjun;School of Electrical Engineering,East China Jiaotong University;School of Electrical Engineering,Beijing Jiaotong University;Beijing Togest Time Automation System Equipment Co.Ltd;

【机构】 华东交通大学电气工程学院； 北京交通大学电气工程学院； 北京太格时代自动化系统设备有限公司；

【摘要】 针对智能配电网量测点多、运行变化快,容易造成海量大数据集信息,使调度产生处理延迟、响应慢的问题,提出了融合变断面压缩存取和分布式集群无损压缩的柔性压缩新算法。综合运用量测点元采样、阀值与时限控制技巧,实现变断面压缩存取,通过构建Hadoop云集群平台,以变断面量测数据作为集群压缩输入对象,利用Map★educe云计算引擎实现分布式无损集群压缩。以西陇海线铁路电调自动化系统为例进行研究,结果表明:变断面压缩法使量测值存取压缩率达到约91%,集群无损压缩法使变断面存储的信息量进一步压缩75%左右,具有显著的柔性技术特征,为智能配电网量测大数据集信息研究提供了一种无损压缩新方法。更多还原

【Abstract】 Massive large data sets measurement information may cause processing delay and slow response in intelligent distribution system dispatching.To solve the problem a new flexible compression algorithm was proposed to deal with massive information,which combined variable section compression with distributed cluster lossless compression.Measurement point sampling,threshold and timing control techniques were used synthetically to achieve variable section compression access,and variable section measurement data were applied to clusters compression input objects in hadoop platform,then Map/Reduce cloud calculation engine was used to achieve distribution cluster lossless compression.Taking Long-Hai line power dispatching automation system as an example,the case study shows that variable section compression method makes access compression rate of measurement value reach about 91%,and clustering lossless compression method makes the amount of variable section stored information further compressed about 75%,flexible technique characteristics are significant,which provide a new lossless compression method for large data sets measurement information of intelligent distribution network.更多还原