摘要
电气设备接地故障辨识方法通常依赖于传感器监测结果,忽视了电气设备间的关联性,导致辨识精度较差。为此,提出基于改进注意力机制的电气设备接地故障自动化辨识方法。分析电气设备的连接关系后,更新电气设备运行数据特征,通过聚合操作实现特征融合。基于此,结合改进注意力机制构建自动化辨识模型,实现对接地故障特征的深度学习与精准分类。实验表明,采用该方法得到的辨识结果与标注标签的一致性较高。
The identification method of grounding faults in electrical equipment usually relies on sensor monitoring results,ignoring the correlation between electrical equipment,resulting in poor identification accuracy.Therefore,this paper proposes an automated identification method for electrical equipment grounding faults based on an improved attention mechanism.After analyzing the connection relationship of electrical equipment,update the operational data characteristics of electrical equipment and achieve feature fusion through aggregation operations.Based on this,combined with an improved attention mechanism,an automated identification model is constructed to achieve deep learning and accurate classification of ground fault features.The experiment shows that the identification results obtained by this method have high consistency with the annotated labels.
作者
王义春
王文琦
WANG Yichun;WANG Wenqi(Boli County Power Supply Branch,State Grid Heilongjiang Electric Power Co.,Ltd.,Qitaihe,Heilongjiang 154600,China;Qitaihe Power Supply Company,State Grid Heilongjiang Electric Power Co.,Ltd.,Qitaihe,Heilongjiang 154600,China)
出处
《自动化应用》
2025年第15期198-200,共3页
Automation Application
关键词
改进注意力机制
电气设备
接地故障
辨识方法
辨识精度
improved attention mechanism
electrical equipment
grounding fault
identification method
identification accuracy
