周阳洋,胡俊华,徐华,尹骏刚,李庆明,吴慧玲
ZHOU Yangyang,HU Junhua,XU Hua,YIN Jungang,LI Qingming,WU Huiling

• 1:国网浙江省电力有限公司
• 2:湖南大学电气与信息工程学院
• 3:湖南湖大华龙电气与信息技术有限公司

摘要(Abstract)：

利用红外热像法检测瓷支柱绝缘子是否存在异常发热，是变电站支柱绝缘子故障诊断的主要方法之一。结合计算机视觉技术提出了一种针对瓷支柱绝缘子红外图像的轻量级目标检测模型。首先，在深度可分离卷积中加入膨胀卷积核，有效增大输出单元的感受野，减少参数量。然后，使用得到的DMobilenet网络结构替换YOLOv7中的主干网络ELANCSP，并采用SJS（剪切、抖动、缩放）方法扩充样本数量，同时引入迁移学习、Mosaic数据增强、余弦退火等算法提高模型泛化能力。最后，将该模型与YOLOv4、YOLOv5、YOLOv7、G-Adaboost目标检测算法进行了性能对比。实验结果表明，该模型在保证准确率和速度的同时，具有更强的鲁棒性和泛化能力，且模型更轻量化。
Detecting abnormal heat in porcelain post insulators using infrared thermal imaging(ITI) stands as a principal approach for diagnosing fault in post insulators within substations. A lightweight target detection model for infrared images of porcelain post insulators is proposed based on computer vision. First, a dilated convolutional kernel is added to the depthwise separable convolution to effectively increase the receptive field of the output unit and reduce the number of parameters. Then, the obtained D-Mobilenet network structure is used to replace the backbone network ELANCSP in YOLOv7, and the SJS(shear, jitter, scale) method is used to expand the number of samples, and algorithms including transfer learning, Mosaic data augmentation, and cosine annealing are introduced to improve the model's generalization ability. Finally, the performance of the model is compared with YOLOv4, YOLOv5, YOLOv7, and the G-Adaboost target detection algorithm. Experimental findings demonstrate that this model boasts a superior combination of lightweight design, robustness, generalization capacity, accuracy, and speed.

关键词(KeyWords)： 瓷支柱绝缘子;红外图像;目标检测;轻量化网络;YOLOv7
porcelain post insulator;infrared image;target detection;lightweight network;YOLOv7

Abstract:

Keywords:

基金项目(Foundation): 国网浙江省电力有限公司科技项目（B311MR220002）

作者(Author): 周阳洋,胡俊华,徐华,尹骏刚,李庆明,吴慧玲
ZHOU Yangyang,HU Junhua,XU Hua,YIN Jungang,LI Qingming,WU Huiling

DOI: 10.19585/j.zjdl.202311010

参考文献(References)：

• [1]焦宗寒，邵鑫明，郑欣，等.基于振动信号频谱高斯混合模型的瓷支柱绝缘子故障诊断[J].电气技术，2021,22(6):36-42.JIAO Zonghan,SHAO Xinming,ZHENG Xin,et al.Fault diagnosis of porcelain post insulator based on Gaussian mixture model of vibration signal spectrum[J]. Electrical Engineering,2021,22(6):36-42.
• [2]佘立伟，欧阳力，何建军.一起10 kV瓷支柱绝缘子异常发热检测分析[J].电工技术，2020(5):63-64.SHE Liwei,OUYANG Li,HE Jianjun.Study of abnormal fever detection of a 10 kV porcelain post insulator[J].Electric Engineering,2020(5):63-64.
• [3]汪佛池，马建桥，律方成，等.污秽和水分对瓷支柱绝缘子发热的影响[J].高电压技术，2015,41(9):3054-3060.WANG Fochi,MA Jianqiao,LüFangcheng,et al.Pollution and water molecule effects on porcelain post insulator heating[J].High Voltage Engineering,2015,41(9):3054-3060.
• [4] WANG A L. Analysis of the research and development status and future development of porcelain insulator fault detection equipment based on a portable insulator detection device[J]. Journal of Physics:Conference Series,2021,1865(2):022054.
• [5]陈重洪，庄建煌，林俊超，等.瓷支柱绝缘子带电状态检测系统的研制[J].高压电器，2020,56(7):212-217.CHEN Chonghong,ZHUANG Jianhuang,LIN Junchao,et al. Development of live detection system for porcelain post insulator[J].High Voltage Apparatus,2020,56(7):212-217.
• [6]王黎明，刘立帅，梅红伟，等.基于脉冲红外热波技术的支柱瓷绝缘子无损检测方法[J].中国电机工程学报，2017,37(24):7359-7366.WANG Liming,LIU Lishuai,MEI Hongwei,et al.Nondestructive testing method of post porcelain insulator based on pulsed infrared thermal wave technology[J]. Proceedings of the CSEE,2017,37(24):7359-7366.
• [7]姚建刚，付鹏，李唐兵，等.基于红外图像的绝缘子串自动提取和状态识别[J].湖南大学学报（自然科学版），2015,42(2):74-80.YAO Jiangang,FU Peng,LI Tangbing,et al. Algorithm research of automatically extracting the area of insulator from infrared image and state identification[J].Journal of Hunan University(Natural Sciences),2015,42(2):74-80.
• [8]陈芳，姚建刚，李佐胜，等.绝缘子串红外图像中单个绝缘子盘面的提取方法[J].电网技术，2010,34(5):220-224.CHEN Fang,YAO Jiangang,LI Zuosheng,et al. The method to extract shed surface image of a single insulator from infrared image of a insulator string[J].Power System Technology,2010,34(5):220-224.
• [9]刘洋，陆倚鹏，高嵩，等.边缘检测在盘形悬式瓷绝缘子串红外图像上的应用[J].电瓷避雷器，2020(1):198-203.LIU Yang,LU Yipeng,GAO Song,et al.Edge detection on infrared image of high voltage porcelain disc type suspension insulator strings[J]. Insulators and Surge Arresters,2020(1):198-203.
• [10]杨高坤.基于改进YOLOv4的电力设备红外图像故障诊断[D].淮南：安徽理工大学，2021.YANG Gaokun. Infrared image fault diagnosis of power equipment based on improved YOLOv4[D].Huainan:Anhui University of Science&Technology,2021.
• [11]段中兴，张雨明，马佳豪.基于改进YOLOv4的电力设备红外图像识别[J].激光与光电子学进展，2022,59(24):29-36.DUAN Zhongxing,ZHANG Yuming,MA Jiahao.Infrared image recognition of power equipment using improved YOLOv4[J].Laser&Optoelectronics Progress,2022,59(24):29-36.
• [12]顾星，詹伟达，崔紫薇，等.基于注意力机制的红外目标检测方法[J].激光与光电子学进展，2023,60(10):293-300.GU Xing,ZHAN Weida,CUI Ziwei,et al.Infrared target detection method based on attention mechanism[J].Laser&Optoelectronics Progress,2023,60(10):293-300.
• [13]刘冬，李庭鑫，杜宇等.基于MCA-YOLO的轻量级红外实时目标检测算法[J/OL].华中科技大学学报（自然科学版）：1-7[2023-03-08].LIU Dong,LI Tingxin,DU Yu,et al.Lightweight infrared real-time target detection algorithm based on MCAYOLO[J/OL]. Journal of Huazhong University of Science and Technology(Natural Science Edition):1-7[2023-03-08].
• [14]王媛彬，李媛媛，段誉，等.基于轻量骨干网络和注意力结构的变电设备红外图像设备[D].电网技术，2022.WANG Yuanbin,LI Yuanyuan,DUAN Yu,et al.Infrared image recognition of substation equipment based on lightweight backbone network and attention mechanism[D].Power System Technology,2022.
• [15]刘国特，伍伟权，郭芳，等.基于改进级联Gentle Adaboost分类器的支柱绝缘子红外图像AI识别[J].高电压技术，2022,48(3):1088-1095.LIU Guote,WU Weiquan,GUO Fang,et al.AI recognition of post insulator infrared image based on improved cascade gentle adaboost classifier[J].High Voltage Engineering,2022,48(3):1088-1095.