[1]徐志宗,刘 洋,胡一凡,等. 基于GDPC-YOLOv7的螺栓和线夹锈蚀检测方法[J].机械与电子,2026,44(02):9-16.
　XU Zhizong,LIU Yang,HU Yifan,et al. Bolt and Clamp Corrosion Detection Method Based on GDPC-YOLOv7 Algorithm[J].Machinery & Electronics,2026,44(02):9-16.
点击复制

 基于GDPC-YOLOv7的螺栓和线夹锈蚀检测方法()
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
44
期数:
2026年02期
页码:
9-16
栏目:
研究与设计
出版日期:
2026-02-26

文章信息/Info

Title:
 Bolt and Clamp Corrosion Detection Method Based on GDPC-YOLOv7 Algorithm
文章编号:
1001-2257(2026)02-0009-08
作者:
 徐志宗1; 2; 刘 洋1; 胡一凡1; 王新良1
 (1.河南理工大学物理与电子信息学院,河南 焦作 454003;
2.鹤壁煤业(集团)有限责任公司供电处,河南 鹤壁 458000)
Author(s):
 XU Zhizong1; 2; LIU Yang1; HU Yifan1; WANG Xinliang1
 (1.School of Physics and Electronic Information Engineering,Henan Polytechnic University,Jiaozuo 454003,China;
2.Power Supply Department,Hebi Coal Industry Co.,Ltd.,Hebi 458000,China)
关键词:
输电线路; 锈蚀检测; YOLOv7; 轻量化; 注意力机制
Keywords:
transmission lines; corrosion detection; YOLOv7; lightweight; attention mechanism
分类号:
TP391.4
文献标志码:
A
摘要:
针对矿区输电线路中螺栓和线夹锈蚀检测背景复杂、检测精度低的问题,提出了一种轻量化的矿区输电线路螺栓和线夹锈蚀检测算法GDPC YOLOv7。首先,构建GS ELAN 模块,通过引入GSBottleneck模块,与原网络的ELAN 模块相结合,轻量化主干网路,降低模型的参数量;其次,设计DW SPPCSPC模块,以深度可分离卷积(DWConv)代替SPPCSPC模块中的标准卷积,保障模型特征提取能力的同时减少模型的参数量;再次,构造P ELAN 模块,对网络中的ELAN H 模块进行剪枝处理,并引入部分卷积PConv,减少模型的参数量和计算开销;最后,在颈部网络中嵌入坐标注意力机制(CA),增强模型对重要特征的关注度,提高模型的检测精度。实验结果表明,GDPC YOLOv7 算法平均检测精度均值达到了91.31%,相较于原网络提高了1.85百分点,参数量从49.25×106 减少至19.69×106,帧率从58.47帧/s增加至68.96帧/s,改进后的模型良好地平衡了检测精度与速度,满足了矿区输电线路中螺栓和线夹锈蚀检测的需求。
Abstract:
 Addressing the issues of complex backgrounds and low detection accuracy in corrosion detection for bolt and clamp on transmission lines in mining areas,this paper proposes a lightweight detection algorithm named GDPC YOLOv7.Firstly,a GS ELAN module is constructed by introducing the GSBottleneck module and integrating it with the original ELAN module,thereby streamlining the backbone network and reducing the model’s parameter count.Secondly,a DW SPPCSPC module is designed,where the standard convolution in the SPPCSPC module is replaced with Depthwise Separable Convolutions (DWConv).This ensures the model’s feature extraction capability while decreasing its parameters.Thirdly,a P ELAN module is constructed by pruning the ELAN H module within the network and introducing Partial Convolution (PConv),which further reduces both the parameter count and computational overhead.Finally,a Coordinate Attention (CA) mechanism is embedded into the neck network to enhance the model’s focus on crucial features and improve detection accuracy.Experimental results demonstrate that the GDPC YOLOv7 algorithm achieves a mean Average Precision (mAP) of 91.3%,representing an improvement of 1.85 percentage points over the original network.The parameter number is reduced from 49.25×106 to 19.69×106,and the frame rate from 58.47 frames/s to 68.96 frames/s.The improved model effectively balances the detection accuracy and speed,meeting the requirements for corrosion detection of bolt and clamps on transmission lines in mining areas.

参考文献/References:

 [1] 刘传洋,吴一全,刘景景.基于视觉的输电线路金具锈蚀缺陷检测方法研究进展[J].仪器仪表学报,2024,45(3):286 305.
[2] 王凌云,李婷宜,李阳,等.基于FEF DeepLabV3+的电力金具锈蚀分割方法[J].电子测量与仪器学报,2023,37(7):166 176.
[3] 蓝贵文,任新月,徐梓睿,等.基于YOLOv8s的轻量级绝缘子多缺陷检测模型[J].现代电子技术,2024,47(20):72 80.
[4] 邓伟,王洪亮.基于级联网络的螺栓锈蚀检测方法研究[J].现代电子技术,2023,46(19):111 115.
[5] HOWARD A,SANDLER M,CHEN B,et al.Searching for mobilenetv3[C]∥2019 IEEE/CVF International Conference on Computer Vision (ICCV),2019:1314 1324.
[6] WOO S,PARK J,LEE J Y,et al.Cbam:convolutional block attention module[C]∥Proceedings of the European Conference on Computer Vision (ECCV),2018:3 19.
[7] 吴军,白梁军,董晓虎,等.基于Cascade R CNN 算法的输电线路小目标缺陷检测方法[J].电网与清洁能源,2022,38(4):19 27.
[8] CAI Z W,VASCONCELOS N.Cascade R CNN:delving into high quality object detection[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2018:6154 6162.
[9] 刘军,孙庆,刘玮,等.基于U Net网络和椭圆度量学习的防震锤锈蚀识别[J].计算机技术与发展,2020,30(11):163 167.
[10] RONNEBERGER O,FISCHER P,BROX T.U net:convolutional networks for biomedical image segmentation[C]∥Medical Image Computing and Computer Assisted Intervention– MICCAI,2015:234 241.
[11] SONG Z W,HUANG X B,JI C,et al.Deformable YOLOX:detection and rust warning method of transmission line connection fittings based on image processing technology[J].IEEE Transactions on instrumentation and measurement,2023,72:1 21.
[12] CHEN X J,AN Z Y,HUANG L S,et al.Surface defect detection of electric power equipment in substation based on improved YOLOV4 algorithm[C]∥2020 10th International Conference on Power and Energy Systems (ICPES).New York:IEEE,2020:256 261.
[13] WANG C Y,BOCHKOVSKIY A,LIAO H Y M.YOLOv7:trainable bag of freebies sets new state of the art for real time object detectors[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2023:7464 7475.
[14] HOU Q B,ZHOU D Q,FENG J S.Coordinate attention for efficient mobile network design[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2021:13713 13722.
[15] HE K M,ZHANG X Y,REN S Q,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J].IEEE Transactions on pattern analysis and machine intelligence,2015,37(9):1904 1916.
[16] CHOLLET F.Xception:deep learning with depthwise separable convolutions[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2017:1251 1258.
[17] CHEN J R,KAO S H,HE H,et al.Run,don’t walk:chasing higher FLOPS for faster neural networks[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2023:12021 12031

相似文献/References:

[1]生红莹,刘 欢,朱 琳,等.带电线路中施工机械安全距离作业下的临界电场研究[J].机械与电子,2021,(01):12.
　SHENG Hongying,LIU Huan,ZHU Lin,et al.Study on the Critical Electric Field of Construction Machinery Working at Safe Distance in Operating Lines[J].Machinery & Electronics,2021,(02):12.
[2]董泽才,刘昌帅,冒文兵.双程摆扫激光测距探测成像在输电线路通道监测中的应用[J].机械与电子,2021,(06):39.
　DONG Zecai,LIU Changshuai,MAO Wenbing.Application of Two-way Swing Scanning Laser Ranging Detection Imaging in Transmission Line Channel Monitoring[J].Machinery & Electronics,2021,(02):39.
[3]王维坤,余志伟,马鹏飞.输电线路巡检智能单兵装备全景采集系统设计[J].机械与电子,2021,(09):32.
　WANG Weikun,YU Zhiwei,MA Pengfei.Design of Panoramic Acquisition System for Intelligent Single Soldier Equipment in Transmission Line Inspection[J].Machinery & Electronics,2021,(02):32.
[4]刘 阳,王英英.基于大数据的输电线路阻抗参数预测方法[J].机械与电子,2021,(10):15.
　LIU Yang,WANG Yingying.Big Data-based Method for Predicting Transmission Line Impedance Parameters[J].Machinery & Electronics,2021,(02):15.
[5]马洲俊,王茂飞,蒋承伶.一种融合直线机制与膨胀算法的输电线路异物检测方法[J].机械与电子,2024,42(05):29.
　MA Zhoujun,WANG Maofei,JIANG Chengling.Detection Method of Foreign Objects on Transmission Lines Based on Linear Detection Mechanism and Regional Expansion Algorithm[J].Machinery & Electronics,2024,42(02):29.
[6]易莹鑫,吴晓鸣,杨国强,等.适用于山区岩石地形的输电线路膨胀地锚优化设计[J].机械与电子,2025,(04):19.
　YI Yingxin,WU Xiaoming,YANG Guoqiang,et al.Optimization Design of Expansion Anchor for Transmission Line Suitable for Mountainous Rock Terrain[J].Machinery & Electronics,2025,(02):19.
[7]吴奇伟,薛 海,王 真,等.基于轻量化深度学习驱动的输电线路异物入侵实时监测方法[J].机械与电子,2025,(12):38.
　WU Qiwe,XUE Hai,WANG Zhen,et al.A Real-time Monitoring Method for Foreign Object Intrusion in Transmission Lines Driven by Lightweight Deep Learning[J].Machinery & Electronics,2025,(02):38.

备注/Memo

备注/Memo:
 收稿日期:2025-09-30
基金项目:河南省高等学校青年骨干教师培养计划(2019GGJS060);河南省高等学校重点科研基金项目(21B413005)
作者简介:徐志宗 (1980-),男,河南鹤壁人,硕士,高级工程师,研究方向为智能电网;刘 洋 (1999-),男,河南焦作人,硕士,研究方向为人工智能技术;胡一凡 (2001-),男,河南汝州人,硕士,研究方向为人工智能技术;王新良 (1980-),男,河南许昌人,博士,教授,研究方向为人工智能、深度学习和软件设计与开发。
更新日期/Last Update: 2026-04-28