基于实测不均衡小样本的配电网高阻接地故障检测方法  

Detection Method of High-impedance Fault in Distribution Network Based on Uneven Small Samples from Actual Measurements

作　　者：高伟[1] 何文秀 郭谋发[1] 白浩 GAO Wei;HE Wenxiu;GUO Moufa;BAI Hao(College of Electrical Engineering and Automation,Fuzhou University,Fuzhou 350108,China;China Southern Power Grid Research Institute Co.,Ltd.,Guangzhou 510663,China)

机构地区：[1]福州大学电气工程与自动化学院,福州350108 [2]南方电网科学研究院有限责任公司,广州510663

出　　处：《高电压技术》2025年第3期1135-1144,I0001,共11页High Voltage Engineering

基　　金：福建省自然科学基金(2021J01633)。

摘　　要：为了应对实际配电网高阻接地故障信号微弱多变、数据稀缺等问题,提出一种基于实测不均衡小样本的高阻接地故障检测新方法。首先,使用基于压缩-激励网络的多头变分自编码器增殖模型,扩充小样本数据集。其次,将数据进行滤波处理后,分别提取其时、频域特征。鉴于高阻故障特征微弱,增殖模型无法生成全面、有效的故障特征这一事实,进一步提出基于梯度调和机制的类别型特征提升(gradient harmonized mechanism-categorical boosting,GHM-Cat Boost)算法,引入梯度调和机制损失函数,让模型均衡易分样本和难分样本的关注度,从而解决过拟合问题。研究结果表明,数据增殖模型能够生成兼具仿真数据多样性与实测数据随机性特点的故障样本,提高了数据的可利用性。且所提GHM-Cat Boost模型的故障识别准确率可以达到97.21%,优于其对比分类器模型。通过测试和对比分析,验证了所提方案的有效性。In order to address the challenges posed by weak and variable high-impedance fault signals and limited data availability in practical distribution networks,a novel method for detecting high-impedance faults is proposed.Initially,a multi-head variational autoencoder model based on squeeze-excitation networks is employed to augment the small sample dataset.Subsequently,the data are filtered,and the temporal and frequency domain features are extracted,respectively.Considering the weak characteristics of high impedance fault features and the limitations of the proliferation model in generating comprehensive and effective fault features,a categorical boosting algorithm based on the gradient harmonized mechanism(GHM-CatBoost)is introduced.The GHM-CatBoost algorithm incorporates a gradient harmonized mecha-nism loss function to address the imbalance in attention between easily distinguishable and challenging samples,thereby mitigating the issue of overfitting.The research findings suggest that the data proliferation model can produce fault sam-ples with a blend of simulation data diversity and measured data randomness,thereby enhancing the richness of the dataset.Furthermore,the fault recognition accuracy achieved by the proposed GHM-CatBoost model is notably high at 97.21%,outperforming its counterpart classifier model.Moreover,the efficacy of the proposed approach is validated through rigorous testing and comparative analysis.

关 键 词：配电网 高阻接地 故障检测 时频特征提取 变分自编码器 注意力机制 CatBoost 

分 类 号：TM862[电气工程—高电压与绝缘技术]