摘要
为构建高效准确的电网能效诊断系统,解决传统单一数据源诊断精度不高的问题,设计了一种三层数据融合架构,集成SCADA系统、智能电表、气象环境等多源异构数据,采用深度学习自编码器建立异常检测模型,构建涵盖电能质量、设备效率、系统损耗、负荷特性的多维度评估体系。结果表明:所提方法在数据处理速度上比传统方法提升104%,融合准确率达94.3%,诊断准确率达93.2%,误警率仅2.3%,证明基于多源数据融合的电网能效诊断方法显著提升了诊断精度和处理效率,为电网智能化运维提供了有效的技术支撑。
To build an efficient and accurate power grid energy efficiency diagnosis system and solve the problem of low diagnostic accuracy of traditional single data sources,the study designs a three-layer data fusion architecture;integrates multi-source heterogeneous data such as SCADA systems,smart electricity meters and meteorological environment and;establishes an anomaly detection model through deep learning autoencoder and a multi-dimensional evaluation system covering power quality,equipment efficiency,system loss and load characteristics.The results show that the proposed method improves the data processing speed by 104%compared with the traditional method,with a fusion accuracy rate of 94.3%,a diagnostic accuracy rate of 93.2%,and a false alarm rate of only 2.3%.This proves that the power grid energy efficiency diagnosis method based on multi-source data fusion significantly improves the diagnostic accuracy and processing efficiency,and provides effective technical support for the intelligent operation and maintenance of the power grid.
作者
吕宗迎
LüZongying(Shandong Agricultural University,Tai’an 271000,China)
出处
《黑龙江科学》
2025年第24期96-99,102,共5页
Heilongjiang Science
关键词
多源数据
融合
电网
能效
诊断
Multi-source data
Integration
Power grid
Energy efficiency
Diagnosis
