摘要
文章提出一种基于深度学习模型的配电室知识图谱构建方法。首先,以配电室各类设备的异构文本数据为对象,训练深度学习模型进行命名实体识别、关系及事件抽取任务,实现异构数据知识的抽取,进而构建包括设备信息及设备故障处置策略在内的配电室知识图谱;其次,利用Neo4j图数据库实现知识图谱的可视化管理,辅助电网调度人员进行设备高效管理及故障处置决策。在实际应用中,配电室知识图谱能够有效提高设备管理的高效性及故障处理决策的效率及效果。
This paper presents a deep learning-based method for constructing a knowledge graph of the power distribution room.First,taking heterogeneous text data of various devices in the power distribution room as objects,the deep learning model is trained to carry out named entity recognition,relationship and event extraction tasks,and realize the extraction of heterogeneous data knowledge,and then the knowledge graph of the power distribution room including equipment information and equipment fault disposal strategies is constructed.The visual management of the knowledge graph is implemented using the Neo4j graph database to assist power grid dispatching personnel in making efficient equipment management and fault-handling decisions.In practical applications,the knowledge graph of the power distribution room significantly enhances the efficiency of equipment management as well as the effectiveness and speed of fault resolution.
作者
刘康
王颖舒
晏冰洋
张光辉
陈子杭
王平
LIU Kang;WANG Yingshu;YAN Bingyang;ZHANG Guanghui;CHEN Zihang;WANG Ping(Power Dispatching and Control Center,Guizhou Power Grid Co.,Ltd,Guizhou,550002,China;Guiyang Power Supply Bureau,Guizhou Power Grid Co.,Ltd,Guizhou,550001;School of Communication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing,400065,China)
出处
《长江信息通信》
2025年第9期99-102,共4页
Changjiang Information & Communications
基金
贵州电网有限责任公司电力调度控制中心2024年“10kV及以下低压地下配电室全状态感知及高效传输技术研究”(0665002024030103TX00023/060000KC23100029)资助。
关键词
命名实体识别
深度学习
知识图谱
知识抽取
Named entity recognition
Deep learning
Knowledge graph
Knowledge extraction
