基于Spark的电力设备在线监测数据可视化方法  

Visualization method of electrical equipment online monitoring data based on Spark

作　　者：曲朝阳[1] 熊泽宇[1] 颜佳[2] 辛鹏[3] 曲楠 

机构地区：[1]东北电力大学信息工程学院,吉林省吉林市132012 [2]国网吉林省电力有限公司,吉林长春130021 [3]国网吉林省电力有限公司吉林供电公司,吉林省吉林市132001 [4]江苏省电力检修分公司,江苏南京210008

出　　处：《电工电能新技术》2016年第11期72-80,共9页Advanced Technology of Electrical Engineering and Energy

基　　金：国家自然科学基金项目(51277023);吉林省科技计划重点项目(20130206085SF);吉林省科技重点转化项目(20140307008GX)

摘　　要：大数据可视化可以实现海量电力设备在线监测数据中各种属性、运行状态等电力特征信息的图形、图像化直观呈现,为设备运行状态的及时有效监控分析提供有力保障。因此,本文提出一种基于Spark的电力设备在线监测数据可视化方法,为实现大数据环境下的电力设备在线监测数据的状态信息快速提取,在Spark大数据计算平台上,建立了基于设备状态评估指标体系与模糊C均值聚类(FCM)的电力设备状态信息提取算法。针对数据的多维、时序特性,构建三维平行散点图的数据可视化展现形式,实现电力设备在线监测数据信息全貌的可视化展现。将该方法运用于吉林省某风电场的风电机组在线监测数据集,实验结果证明了该方法的有效性。With the strengthening of range and quality of the electrical equipment online monitoring in smart grid,the collected data volume in online monitoring is growing exponentially. All the attributes and the operating state in electrical equipment online monitoring data which is of massive amounts can be presented directly by the big data visualization,which can provide powerful guarantee to effective and timely monitoring and analyzing of the operating state. However,the application of the big data visualization to electrical power big data is still in a preliminary stage,and there is still lacking of visualization method of electrical equipment online monitoring data under big data environment. Therefore,a visualization method of the electrical equipment online monitoring data based on Spark is proposed. To realize rapid extraction of electrical equipment state information,the state information extracting algorithm based on FCM and evaluation system is constructed on Spark. For the multi-dimensional and time-serial feature of electrical equipment online monitoring data,the representation based on 3-D parallel scatter is constructed,and the data information of the electrical equipment online monitoring is realized. The method is applied to the WTGS online monitoring data set,and the efficiency of the method is proved by the experiment result.

关 键 词：SPARK 电力大数据 信息可视化 在线监测 

分 类 号：TM46[电气工程—电器]