基于CWT-CBAM-CNN的配网接地故障选线方法  

Distribution Network Grounding Fault Line Selection Method Based on CWT-CBAM-CNN

作　　者：胡祥谢 聂祥论 谢宪源 Xiangxie Hu;Xianglun Nie;Xianyuan Xie(School of Electrical Engineering,Guizhou University,Guiyang Guizhou;Bijie Power Supply Bureau of Guizhou Power Grid Co.,Ltd.,Bijie Guizhou;Guizhou Wujiang Qingshuihe Hydropower Development Co.,Ltd.,Guiyang Guizhou)

机构地区：[1]贵州大学电气工程学院,贵州贵阳 [2]贵州电网有限责任公司毕节供电局,贵州毕节 [3]贵州乌江清水河水电开发有限公司,贵州贵阳

出　　处：《建模与仿真》2025年第1期31-38,共8页Modeling and Simulation

摘　　要：针对配网单相接地故障特征提取困难,且现有选线方法选线精度不高的问题,提出了一种连续小波变换(continuous wavelet transform,CWT)和CBAM-CNN的故障选线方法。首先,对零序暂态电流进行连续小波变换获取对应的时频灰度图像;然后,构建了融合CBAM的CNN故障检测模型,使模型更加关注有效信息。仿真结果表明,所提方法的选线精度为99.45%,与CNN相比,具有较强的鲁棒性。Aiming at the problem that the fault feature extraction of single-phase grounding fault in distribution network is difficult and the accuracy of fault line selection is not high,a fault line selection method based on continuous wavelet transform(CWT)and CBAM-CNN is proposed.Firstly,the corresponding time-frequency gray image is obtained by continuous wavelet transform of zero-sequence transient current.Then,a CNN fault phase detection model with CBAM is constructed,which makes the model pay more attention to effective information.The simulation results show that the line selection accuracy of the proposed method is 99.45%.Compared with CNN,it has strong robustness.

关 键 词：故障选线 连续小波变换 卷积神经网络 特征提取 

分 类 号：TM862[电气工程—高电压与绝缘技术]