天津城南变电站多源异构数据的知识图谱构建

A Knowledge Graph Construction Method of Multi-source Heterogeneous Data for Tianjin Chengnan Power Plant Station

新一代电力系统的建设与发展使得变电站及其设备之间的关系复杂化,运行产生的数据也呈现出多源异构的特点。为了得到关联所有复杂设备与多元数据的全域模型,一种针对天津城南变电站数据的知识图谱构建方法在此被提出。首先对来自变电站的多源异构数据进行特征分析与知识抽取,并形成电力系统实体关系库和电力系统故障信息库。知识抽取的过程包括语句分词、关键词抽取、图像转文本等操作。其次设计出基于天津城南变电站的知识图谱结构,并构建出涵盖厂站与厂站、厂站与设备、设备与设备之间实体关系的知识图谱,形成一体化数据平台。最后简单分析了该知识图谱的应用。结果表明,变电站知识图谱可以涵盖变电站所有设备,并完成大量运行数据的存储与关联。

The relationship between power substations and their equipment has been complicated by the new generation of power systems,whose data presents the characteristics of multi-source and heterogeneity.To obtain a global model,a knowledge graph construction method for Tianjin Chengnan power substation is proposed.Firstly,the feature analysis and knowledge extraction of multi-source and heterogeneous data are carried out,which is adopted to form the entity relation database and fault information database of power system.Secondly,the knowledge graph structure based on Tianjin Chengnan Power substation is designed.The knowledge graph is constructed to form an integrated data platform,which covers the entity relationship between station and equipment.Finally,the application of the knowledge graph is briefly analyzed.The results show that the knowledge graph of power substation can cover all the equipment and large data.