基于多源数据融合的智能配电网变压器健康状态在线监测方法  

Online Monitoring Method for the Health Status of Transformers in Intelligent Distribution Networks Based on Multi-Source Data Fusion

作　　者：綦俊 QI Jun(The Fifth Thermal Power Co.,Ltd.,of Qingdao Energy&Thermal Power Group,Qingdao 266041,China)

机构地区：[1]青岛能源热电集团第五热力有限公司,山东青岛266041

出　　处：《通信电源技术》2025年第4期246-248,共3页Telecom Power Technology

摘　　要：变压器健康状态对电网的稳定性至关重要。传统变压器健康监测方法依赖单一信号源或定期检查,难以实现全面、实时的监控。文章提出基于多源数据融合的智能配电网变压器健康状态在线监测方法,设计小波变换采集多源数据,引入谱熵与希尔伯特变换提取健康状态特征,应用加权融合技术组合多源数据特征,借助机器学习模型评估健康状态。实验结果表明,文章方法在准确率、计算复杂度、特征提取时间以及评估时间等方面均优于传统方法,支持智能电网中变压器的实时监控与健康管理。The health status of transformers is crucial for the stability of the power grid.Traditional transformer health monitoring methods rely on a single signal source or regular inspections,making it difficult to achieve comprehensive and real-time monitoring.This paper propose an online monitoring method of transformer health status in smart distribution network based on multi-source data fusion is proposed,wavelet transform is designed to collect multi-source data,spectral entropy and Hilbert transform are introduced to extract health status features,and multi-source data features are combined by weighted fusion technology,and health status is evaluated by machine learning model.The experimental results show that the proposed method outperforms traditional methods in terms of accuracy,computational complexity,feature extraction time,and evaluation time,effectively supporting real-time monitoring and health management of transformers in smart grids.

关 键 词：智能配电网 变压器健康状态 多源数据融合 在线监测 健康评估 

分 类 号：TM41[电气工程—电器]