基于电网多源数据的电力知识图谱构建方法  

Construction method of power knowledge map based on multi⁃source data of power grid

作　　者：张凡 陈浩敏 姚森敬 邓远发 ZHANG Fan;CHEN Haomin;YAO Senjing;DENG Yuanfa(China Southern Power Grid Digital Grid Research Institute Co.,Ltd.,Guangzhou 510700,China;China Southern Power Grid Shenzhen Digital Grid Research Institute Co.,Ltd.,Shenzhen 518000,China)

机构地区：[1]南方电网数字电网研究院有限公司,广东广州510700 [2]南方电网深圳数字电网研究院有限公司,广东深圳518000

出　　处：《电子设计工程》2024年第9期175-178,183,共5页Electronic Design Engineering

摘　　要：电力知识图谱构建方法重复采集电网多源数据,针对构建得到的知识图谱置信度过小的问题,设计一种基于电网多源数据的电力知识图谱构建方法。通过设定实体电网多源数据采集架构,采用局部线性嵌入算法筛选重复采集的电网多源数据。将电力知识间的关系转变可识别特征数值,抽取多源数据中的电力知识划分图谱层级,构建电力知识图谱结构。结合电力知识采集环境完成实验,实验结果表明,文中设计的构建方法准确率为97%,召回率为0.78%,F值为93,满足电力知识图谱的构建需求。Power knowledge map construction method repeatedly collects multi-source data of power grid.Aiming at the problem that the confidence of the constructed knowledge map is too low,a power knowledge map construction method based on multi-source data of power grid is designed.By setting the multi-source data acquisition architecture of the physical power grid,the local linear embedding algorithm is used to screen the repeated multi-source data of the power grid.The relationship between power knowledge is transformed into identifiable characteristic values,and the power knowledge from multi-source data is extracted to divide the map hierarchy,and the power knowledge map structure is constructed.Combined with the electric power knowledge collection environment,the experiment results show that the accuracy rate of the construction method designed in this paper is 97%,the recall rate is 0.78%,and the F value is 93,which meets the construction requirements of electric power knowledge map.

关 键 词：电网多源数据 知识图谱 局部线性嵌入算法 逻辑参数 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]