Electrical Contact Defect Identification of GIS Bus Based on Current-magnetic Field Data Fusion and Cross-attention

XIE Zhiyang; PENG Tao; JING Lei

doi:10.13296/j.1001-1609.hva.2025.10.013

您当前的位置：

首页 >

文章列表页 >

Electrical Contact Defect Identification of GIS Bus Based on Current-magnetic Field Data Fusion and Cross-attention

Digital Twin Technology and Applications for Multi-scenario in New Power System | 更新时间：2025-10-21

- Electrical Contact Defect Identification of GIS Bus Based on Current-magnetic Field Data Fusion and Cross-attention
- High Voltage Apparatus Vol. 61, Issue 10, Pages: 117-127(2025)
- 作者机构：
- 作者简介：
- 基金信息：
- DOI：10.13296/j.1001-1609.hva.2025.10.013
  CLC：
- Received：27 March 2025，
  
  Revised：2025-07-02，
  
  Published：16 October 2025
- 稿件说明：
移动端阅览
XIE Zhiyang, PENG Tao, JING Lei. Electrical Contact Defect Identification of GIS Bus Based on Current-magnetic Field Data Fusion and Cross-attention[J]. High voltage apparatus, 2025, 61(10): 117-127.
DOI：

XIE Zhiyang, PENG Tao, JING Lei. Electrical Contact Defect Identification of GIS Bus Based on Current-magnetic Field Data Fusion and Cross-attention[J]. High voltage apparatus, 2025, 61(10): 117-127. DOI： 10.13296/j.1001-1609.hva.2025.10.013.

摘要

气体绝缘组合电器（GIS）母线触头的对接深度不足、对中度偏差和弹簧松弛等接触缺陷会显著增加接触电阻，引发过热甚至绝缘击穿事故。针对现有检测方法在复杂工况下灵敏度不足的问题，文中提出一种基于电流—磁场双模态数据融合与交叉注意力机制的缺陷识别算法。通过GIS原型试验平台同步采集4种典型工况下的瞬态电流信号与16通道空间磁场分布数据，构建多物理场数据集；采用Mel滤波器提取电流时频特征，通过对称点模式（SDP）表征磁场空间特性；进而设计双分支ResNet-18网络分别处理两类特征，通过交叉注意力机制动态分配模态权重，实现1 024维融合特征的深度提取；最终经多层感知机分类器输出缺陷类型。实验表明：所提模型在10折交叉验证中平均分类准确率达92.76%，较传统拼接融合和双线性池化方法显著提升。t-SNE可视化证实融合特征具有更优的类间分离度，单次推理耗时低于50 ms，满足工程实时性需求，为GIS电接触状态的在线监测与早期预警提供了高精度解决方案。

Abstract

Such contact defects as insufficient butt depth

alignment deviation and spring relaxion of bus contact of gas insulated switchgear (GIS) will significantly increase contact resistance

leading to overheating and even insulation breakdown accidents. Aiming at the problem of insufficient sensitivity of existing detection methods under complex working conditions

a defect identification algorithm based on current-magnetic field bimodal data fusion and cross-attention mechanism is proposed in this paper. Through GIS prototype test platform

the transient current signal and 16-channel spatial magnetic field distribution data under four typical working conditions are synchronously collected

and a multi-physics dataset is constructed. Mel filter is used to extract the time-frequency characteristics of current

and the spatial characteristics of magnetic field are characterized by symmetric point mode (SDP) . Then

a dual-branch ResNet-18 network is designed to process two types of features respectively

and modal weights are dynamically allocated through cross-attention mechanism to achieve deep extraction of 1 024-dimensional fusion features. Finally

the defect type is output by the multi-layer perceptron classifier. Experiments show that the average classification accuracy of the proposed model in 10-fold cross-validation is 92.76%

which has significant improvement compared with the traditional stitching and fusion and bilinear pooling methods.The t-SNE visualization proves that the fusion features have better inter-class separation

and the single reasoning time is less than 50 ms

which meets the real-time requirements of engineering and provides a high-precision solution for online monitoring and early warning of electrical contact status of GIS.

关键词

Keywords

references

RUNDE M , LILLEVIK O , LARSEN V , et al . Condition assessment of contacts in gas-insulated substations [J ] . IEEE Transactions on Power Delivery , 2004 , 19 ( 2 ): 609 - 617 .

李雷 , 罗容波 , 王俊波 , 等 . 基于千安级冲击电流的GIS电接触特性状态评估研究 [J ] . 高压电器 , 2015 , 51 ( 9 ): 49 - 54 .

LI Lei , LUO Rongbo , WANG Junbo , et al . Assessment of GIS electrical contact state by means of kiloampere grade impact current [J ] . High Voltage Apparatus , 2015 , 51 ( 9 ): 49 - 54 .

张涛 . 频率对GIS稳态温升特性影响的仿真研究 [J ] . 电工电气 , 2022 ( 8 ): 36 - 40 .

ZHANG Tao . Simulation research on the inlfuence of the frequency on characteristics of the stable temperature rise of GIS [J ] . Electrotechnics Electric , 2022 ( 8 ): 36 - 40 .

王雪 . GIS温升特性多物理场耦合仿真与试验研究 [D ] . 沈阳 : 沈阳工业大学 , 2024 .

WANG Xue . Multi-physics coupling simulation and experimental research on GIS temperature rise characteristics [D ] . Shenyang : Shenyang University of Technology , 2024 .

金虎 , 彭在兴 , 王颂 , 等 . GIS触头过热故障实验及外壳红外热像分布特征研究 [J ] . 智慧电力 , 2017 , 45 ( 11 ): 37 - 41 .

JIN Hu , PENG Zaixing , WANG Song , et al . Study on overheat experiment of GIS contact and infrared thermal image distribution on its shell [J ] . Smart Power , 2017 , 45 ( 11 ): 37 - 41 .

丛浩熹 , 李庆民 , 齐波 , 等 . 基于红外传感的GIS隔离开关触头温度在线监测技术研究 [J ] . 电力自动化设备 , 2014 , 34 ( 3 ): 144 - 148 .

CONG Haoxi , LI Qingmin , QI Bo , et al . Online GIS switch contact temperature monitoring based on IR sensing [J ] . Electric Power Automation Equipment , 2014 , 34 ( 3 ): 144 - 148 .

陈强 , 李庆民 , 丛浩熹 , 等 . 基于多点分布式光纤光栅的GIS隔离开关触头温度在线监测技术 [J ] . 电工技术学报 , 2015 , 30 ( 12 ): 298 - 306 .

CHEN Qiang , LI Qingmin , CONG Haoxi , et al . On - line temperature monitoring for GIS disconnecting switch contacts based on multipoint-distributed fiber bragg grating [J ] . Transactions of China Electrotechnical Society , 2015 , 30 ( 12 ): 298 - 306 .

李秀广 , 吴旭涛 , 朱洪波 , 等 . 基于振动信号的GIS触头接触异常研究分析 [J ] . 高压电器 , 2016 , 52 ( 10 ): 165 - 169 .

LI Xiuguang , WU Xutao , ZHU Hongbo , et al . Research on contacting fault of GIS contact based on vibration signal [J ] . High Voltage Apparatus , 2016 , 52 ( 10 ): 165 - 169 .

郭碧红 . 气体绝缘封闭电器内部故障振动法带电监测装置 [J ] . 电工技术杂志 , 1992 ( 3 ): 46 - 47 .

GUO Bihong . Live monitoring device for internal fault vibration method of gas insulated closed electrical appliances [J ] . Journal of Electrical Technology , 1992 ( 3 ): 46 - 47 .

SHINKAI H , GOSHIMA H , YASHIMA M . A study on condition assessment method of gas-insulated switchgear.part II.influence of moisture in the SF 6 , detection of a partial discharge on a spacer, repetition discharge and overheating by incomplete contact [J ] . Electrical Engineering in Japan , 2011 , 176 ( 2 ): 174 - 182 .

何宁辉 , 李秀广 , 周秀 , 等 . GIS固体绝缘不同类型气隙缺陷的放电特征 [J ] . 高电压技术 , 2021 , 47 ( 6 ): 2073 - 2083 .

HE Ninghui , LI Xiuguang , ZHOU Xiu , et al . Discharge characteristics of different types of void defects of solid insulation of GIS [J ] . High Voltage Engineering , 2021 , 47 ( 6 ): 2073 - 2083 .

李博阳 , 廖柯熹 , 何国玺 , 等 . 弱磁检测技术研究进展与展望 [J ] . 电子测量与仪器学报 , 2022 , 36 ( 3 ): 1 - 14 .

LI Boyang , LIAO Kexi , HE Guoxi , et al . Research progresses and prospects of weak magnetic testing technology [J ] . Journal of Electronic Measurement and Instrumentation , 2022 , 36 ( 3 ): 1 - 14 .

王明宇 . 基于磁场信号的永磁同步电机匝间短路和偏心故障诊断研究 [D ] . 哈尔滨 : 哈尔滨工业大学 , 2022 .

WANG Mingyu . Research on turn-to-turn short circuit and eccentric fault diagnosis of permanent magnet synchronous motor based on magnetic field signal [D ] . Harbin : Harbin Institute of Technology , 2022 .

LI Lijun , ZHANG Xiaoguang . Discrimination method of wire rope fault signal based on Holzer sensor for multi array weak magnetic detection [J ] . Cluster Computing , 2019 , 22 ( 2 ): 4901 - 4906 .

FUJINAMI H , TAKUMA T , KAWAMOTO T . Development of detection method with a magnetic field sensor for incomplete contact in gas insulated switches and bus connecting parts [J ] . IEEE Transactions on Power Delivery , 1995 , 10 ( 1 ): 229 - 236 .

程永健 . 基于双轴隧道磁阻传感阵列的风机雷电流反演研究 [D ] . 合肥 : 合肥工业大学 , 2021 .

CHENG Yongjian . Research on fan lightning current inversion based on biaxial tunnel magnetoresistive sensing array [D ] . Hefei : Hefei University of Technology , 2021 .

RAMACHANDRAM D , TAYLOR G W . Deep multimodal learning:A survey on recent advances and trends [J ] . IEEE Signal Processing Magazine , 2017 , 34 ( 6 ): 96 - 108 .

BAI Tangbo , YANG Jianwei , YAO Dechen , et al . Information fusion of infrared images and vibration signals for coupling fault diagnosis of rotating machinery [J ] . Shock and Vibration , 2021 ( 1 ): 6622041 .

吴迪 . 融合视频图像和雷达点云数据的自动驾驶环境信息感知研究 [D ] . 北京 : 北京交通大学 , 2023 .

WU Di . Research on autonomous driving environment information perception by fusing video images and radar point cloud data [D ] . Beijing : Beijing Jiaotong University , 2023 .

高锋阳 , 刘庆寅 , 赵丽丽 , 等 . 改进灰狼优化算法优化CNNLSTM的PEMFC性能衰退预测 [J ] . 电力系统保护与控制 , 2025 , 53 ( 13 ): 175 - 187 .

GAO Fengyang , LIU Qingyin , ZHAO Lili , et al . PEMFC performance degradation prediction based on optimized CNNLSTM using improved grey wolf optimization algorithm [J ] . Power System Protection and Control , 2025 , 53 ( 13 ): 175 - 187 .

胥婧雯 , 于红 , 张鹏 , 等 . 基于声音与视觉特征多级融合的鱼类行为识别模型U-FusionNet-ResNet50+SENet [J ] . 大连海洋大学学报 , 2023 , 38 ( 2 ): 348 - 356 .

XU Jingwen , YU Hong , ZHANG Peng , et al . A fish behavior recognition model based on multi-level fusion of sound and vision U-FusionNet-ResNet50+SENet [J ] . Journal of Dalian Fisheries University , 2023 , 38 ( 2 ): 348 - 356 .

杨晨 , 侯志强 , 李新月 , 等 . 基于CNN-Transformer双模态特征融合的目标检测算法 [J ] . 光子学报 , 2024 , 53 ( 3 ): 273 - 286 .

YANG Chen , HOU Zhiqiang , LI Xinyue , et al . Object detection algorithm based on CNN-Transformer dual modal feature fusion [J ] . Acta Photonica Sinica , 2024 , 53 ( 3 ): 273 - 286 .

谭贵生 , 赵波 , 张桂莲 , 等 . 多策略改进DBO算法与KELM的变压器故障辨识 [J ] . 电力系统保护与控制 , 2025 , 53 ( 14 ): 111 - 122 .

TAN Guisheng , ZHAO Bo , ZHANG Guilian , et al . Transformer fault identification based on multi - strategy improved DBO algorithm and KELM [J ] . Power System Protection and Control , 2025 , 53 ( 14 ): 111 - 122 .

杨景刚 , 刘媛 , 宋思齐 , 等 . GIS设备机械缺陷的振动检测技术研究 [J ] . 高压电器 , 2018 , 54 ( 11 ): 86 - 90 .

YANG Jinggang , LIU Yuan , SONG Siqi , et al . Research on vibration detection technology of GIS mechanical defects [J ] . High Voltage Apparatus , 2018 , 54 ( 11 ): 86 - 90 .

刘媛 , 杨景刚 , 贾勇勇 , 等 . 基于振动原理的GIS隔离开关触头接触状态检测技术 [J ] . 高电压技术 , 2019 , 45 ( 5 ): 1591 - 1599 .

LIU Yuan , YANG Jinggang , JIA Yongyong , et al . Connection state diagnosis method of GIS disconnector based on mechanical vibration [J ] . High Voltage Engineering , 2019 , 45 ( 5 ): 1591 - 1599 .

WU Dehui , SU Lingxin , WANG Xiaohong , et al . A novel nondestructive testing method by measuring the change rate of magnetic flux leakage [J ] . Journal of Nondestructive Evaluation , 2017 , 36 ( 2 ): 107 - 118 .

MA Yilai , HE Renyang , CHEN Jinzhong . A method for improving snr of drill pipe leakage flux testing signals by means of magnetic concentrating effect [J ] . IEEE Transactions on Magnetics , 2015 , 51 ( 9 ): 1 - 7 .

王伟 . 高压输电导线钢芯断股漏磁场矢量计算与分布规律研究 [D ] . 北京 : 华北电力大学 , 2023 .

WANG Wei . Research on calculation and distribution law of leakage magnetic field vector of steel core broken strands of highvoltage transmission conductors [D ] . Beijing : North China Electric Power University , 2023 .

冯搏 , 伍剑波 , 邱公喆 , 等 . 高速漏磁检测方法的发展 [J ] . 无损检测 , 2021 , 43 ( 2 ): 57 - 63 .

FENG Bo , WU Jianbo , QIU Gongzhe , et al . Development of highspeed magnetic flux leakage testing method [J ] . Non-destructive Testing , 2021 , 43 ( 2 ): 57 - 63 .

WU Jianbo , SUN Yanhua , FENG Bo , et al . The effect of motioninduced eddy current on circumferential magnetization in MFL testing for a steel pipe [J ] . IEEE Transactions on Magnetics , 2017 , 53 ( 7 ): 1 - 6 .

杨理践 , 耿浩 , 高松巍 , 等 . 高速漏磁检测饱和场建立过程及影响因素研究 [J ] . 仪器仪表学报 , 2019 , 40 ( 10 ): 1 - 9 .

YANG Lijian , GENG Hao , GAO Songwei , et al . Study on the establishment process and influence factors of high-speed magnetic flux leakage testing [J ] . Chinese Journal of Scientific Instrument , 2019 , 40 ( 10 ): 1 - 9 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Fault Diagnosis Method for DC Microgrids Based on PLO-SVMD and CNN-BiGRU-ATT

Research on Application of Improved YOLOv11n in Detection of Main Defects of Insulators on Transmission Lines

Intelligent Checking Method of Abnormal Electricity Data Based on Stacking Strategy and Deep Learning

Seismic Damage Identification Method for Converter Station Equipment Based on Long-short-term Memory Network

Research on Status Identification of Disconnector Based on ResNet Deep Learning Network

Related Author

HE Shan

BI Xinyin

WANG Kai

QIU Bin

MA Yue

LIU Hongzhi

ZHENG Yingkai

YANG Liu

Related Institution

新疆大学可再生能源发电与并网控制教育部工程研究中心，乌鲁木齐 830017Engineering Research Center of Renewable Energy Power Generation and Grid-connected Control, Ministry of Education, Xinjiang University, Urumqi 830017, China

新疆大学可再生能源发电与并网技术自治区重点实验室，乌鲁木齐 830017Renewable Energy Power Generation and Grid-connected Technology in the Autonomous Region, Xinjiang University, Urumqi 830017, China

国家电网葫芦岛供电公司，辽宁葫芦岛 125000State Grid Huludao Power Supply Company, Liaoning Huludao 125000, China

辽宁工程技术大学电气与控制工程学院，辽宁葫芦岛 125105Faculty of Electrical and Control Engineering, Liaoning Technical University, Liaoning Huludao 125105, China

School of Electric Power Engineering, South China University of Technology

⁰