首页|基于知识图谱的变压器匝间短路故障辨识研究

基于知识图谱的变压器匝间短路故障辨识研究

扫码查看
变压器出现的故障数据间的关联性没有被较好地利用,会直接影响变压器匝间短路故障的辨识准确性.为此,提出基于知识图谱的变压器匝间短路故障辨识研究.基于柔性策略采集变压器数据,根据实际运行情况及数据采集目标需求,调整数据采集量及时间间隔.根据采集到的变压器数据,采用本体构建、实体抽取、关系抽取和图谱构建的步骤构建知识图谱.将提取的知识图谱故障样本特征,输入到你只看一次(YOLO)v4 检测模型中.通过YOLOv4 检测模型与知识图谱结合的检测方法,完成变压器匝间短路故障的自动辨识.试验结果表明:变压器匝间短路故障自动辨识的准确率、召回率和F值均较高,因而辨别及时性高、自动辨别效果好.该研究解决了传统方法中存在的问题,具有重要的现实意义.
Research on Transformer Inter-Turn Short Circuit Fault Identification Based on Knowledge Map
The correlation between the fault data appearing in the transformer is not well utilized,which will directly affect the accuracy of transformer inter-turn short circuit fault identification.For this reason,research on transformer inter-turn short circuit fault identification based on knowledge map is proposed.Based on the flexible strategy to collect transformer data,the amount of data collection and time interval are adjusted according to the actual operating conditions and data collection target demand.Based on the collected transformer data,the knowledge map is constructed using the steps of ontology construction,entity extraction,relationship extraction and map construction.The extracted knowledge map fault sample features are input into the you only look once(YOLO)v4 detection model.The automatic identification of transformer inter-turn short circuit faults is accomplished by the detection method combining the YOLOv4 detection model and the knowledge map.The test results show that the accuracy,recall and F-value of automatic transformer inter-turn short circuit fault identification are high,so the identification timeliness is high,and the automatic identification effect is good.The research solves the problems existing in the traditional method,and has important practical significance.

Knowledge mapTransformerInter-turn short circuit faultsEntity extractionYou only look once(YOLO)v4 detection modelFlexible strategyRelational extraction

查易艺、王翀、张明明

展开 >

国网江苏省电力有限公司信息通信分公司,江苏 南京 210000

知识图谱 变压器 匝间短路故障 实体抽取 你只看一次v4检测模型 柔性策略 关系抽取

2024

自动化仪表
中国仪器仪表学会 上海工业自动化仪表研究院

自动化仪表

CSTPCD
影响因子:0.655
ISSN:1000-0380
年,卷(期):2024.45(4)
  • 12