基于知识图谱的电力动态数据相似性检索方法  

Similarity Retrieval Method for Power Dynamic Data Based on Knowledge Graph

作　　者：李翔 吴少平 刘青 段亚定 王川江 张亮 LI Xiang;WU Shaoping;LIU Qing;DUAN Yading;WANG Chuanjiang;ZHANG Liang(State Grid Corporation of China,Beijing,100031,China;State Grid Info-Telecom Great Power Science and Technology Co.Ltd,Fuzhou,350003,China)

机构地区：[1]国家电网有限公司,北京100031 [2]国网信通亿力科技有限责任公司,福州350003

出　　处：《网络新媒体技术》2025年第2期56-60,共5页Network New Media Technology

摘　　要：为了提升电力动态数据相似性检索准确性与效率,提出基于知识图谱的电力动态数据相似性检索方法。通过构建电力动态数据知识图谱,分析电力系统中不同数据源的实体、属性和关系信息。利用邻接矩阵更新计算矩阵内中心节点与其他节点的杰卡德相似距离,输出电力动态数据相似性检索结果。实验结果表明,所提方法可以有效对电力动态数据进行实时检索,及时发现出现偏差的时序数据节点,且检索准确率最低值为97%,检索效率最高值为98.2%。In order to improve the accuracy and efficiency of similarity retrieval of power dynamic data,a similarity retrieval method of power dynamic data based on knowledge graph is proposed.By constructing the knowledge graph of power dynamic data,the entity,attribute and relationship information of different data sources in power system are analyzed.The adjacency matrix is used to update the Jaccard similarity distance between the central node and other nodes in the calculation matrix,and the similarity retrieval results of power dynamic data are output.The experimental results show that the proposed method can effectively search the power dynamic data in real time,and find the time series data nodes with deviation in time,and the lowest retrieval accuracy is 97%,and the highest retrieval efficiency is 98.2%.

关 键 词：知识图谱 电力动态数据 邻接矩阵 杰卡德相似距离 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]