太赫兹科学与电子信息学报2025，Vol.23Issue(10)：1060-1066,1073,8.DOI:10.11805/TKYDA2023358
基于时序知识图谱的工业物联网设备故障态势预测
Fault situation prediction technology for industrial IoT equipment based on Temporal Knowledge Graph
摘要
Abstract
With the development of the industrial Internet of Things(IoT),industrial systems have become increasingly complex,and the volume of operation and maintenance data from industrial equipment has grown substantially.This makes situational forecasting for industrial equipment especially critical.Knowledge graphs have proven to offer significant advantages in handling large-scale heterogeneous data and can be applied to process operation and maintenance data from industrial IoT devices.Addressing the time-series logical problem of predicting failure trends in industrial equipment,this paper proposes a Temporal Knowledge Graph(TKG)reasoning and representation learning model.The model employs a local recurrent graph encoder network to model historical dependencies of events at adjacent time points,and a global historical encoder network to capture recurring historical facts.Experiments demonstrate that the proposed model outperforms baseline reasoning methods in metrics such as Mean Reciprocal Rank(MRR).关键词
时序知识图谱(TKG)/工业物联网(IoT)/态势预测Key words
Temporal Knowledge Graph(TKG)/industrial Internet of Things(IoT)/situation prediction分类
信息技术与安全科学引用本文复制引用
李世豪,曾锃,缪巍巍,夏元轶,杨君中,沈鹏,孙金龙,赵海涛..基于时序知识图谱的工业物联网设备故障态势预测[J].太赫兹科学与电子信息学报,2025,23(10):1060-1066,1073,8.基金项目
国网江苏省电力有限公司科技资助项目(J2023050) （J2023050）
