摘要
针对传统方法对输变电工程数据分块时间较长,分块准确性差等问题,引入多信息集成技术,设计了一种新的输变电工程数据智能分块方法。通过数据挖掘、数据处理和数据分块3项技术实现分块处理。采用数据挖掘方法对输变电工程数据特性进行偏差分析;依据分析结果设计数据管理流程,通过数据处理对输变电工程数据进行编码分组;根据K最近邻值分类算法进行同类项数据分块处理,实现输变电工程数据的智能分块。为验证分块方法的效果,设计对比实验。结果表明,当输变电工程数据大小为2 000 MB时,设计方法分块时间仅为17 s,较传统方法节省19 s,同时设计分块方法的准确性更高。
The power transmission and transformation project has a large amount of data,and the data is disorderly and disorderly,which requires block processing.However,the existing blocking method has a long processing time and it is not accurate.To this end,the data intelligent block method for power transmission and transformation engineering under multi-information integration is proposed.Obtaining data characteristics of power transmission and transformation engineering through data mining technology,and data feature deviation analysis is performed.Data conversion is used to obtain the true mapping of the database,design the data management process,and encode the data of the power transmission and transformation engineering data.According to the K nearest neighbor value classification algorithm,the same item data block processing is performed to realize intelligent block of data of power transmission and transformation engineering.To verify the effectiveness of the blocking method,a comparative experiment was designed.The results show that when the data size of the power transmission and transformation project is 2000 MB,the design method is only 17 s,which saves 19 s compared with the traditional method,and the accuracy of the design block method is higher.
作者
韩文军
吴红利
薄鑫
许可
HAN Wenjun;WU Hongli;BO Xin;XU Ke(State Grid Economic And Technological Research Institute Co.,Ltd,,Beijing 102209,China;Jiangxi Booway New Technology Co.,Ltd.,Nanchang 330096,China;State Grid Jiangsu Electric Company LimitediNanjing 210024,China)
出处
《自动化与仪器仪表》
2020年第9期204-207,共4页
Automation & Instrumentation
基金
国家电网有限公司总部科技项目资助(No.B3441617K005)。
关键词
多信息集成
输变电工程
智能分块
数据管理流程
multi-information integration
power transmission and transformation engineering
intelligent partitioning
data management process