[1]蔡宇翔,陈丽娟,安　琪.边缘设备端轻量级SSD变电站缺陷检测算法[J].郑州大学学报(工学版),2026,47(01):140-146.[doi:10.13705/j.issn.1671-6833.2025.04.023]
　CAI Yuxiang,CHEN Lijuan,AN Qi.Lightweight SSD Substation Defect Detection Algorithm on the Edge Device Side[J].Journal of Zhengzhou University (Engineering Science),2026,47(01):140-146.[doi:10.13705/j.issn.1671-6833.2025.04.023]
点击复制

边缘设备端轻量级SSD变电站缺陷检测算法()
分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
47
期数:
2026年01期
页码:
140-146
栏目:
出版日期:
2026-01-06

文章信息/Info

Title:
Lightweight SSD Substation Defect Detection Algorithm on the Edge Device Side
文章编号:
1671-6833(2026)01-0140-07
作者:
蔡宇翔1; 2;  陈丽娟3;  安　琪4
1.上海交通大学 电子信息与电气工程学院,上海 200240;2.国网福建省电力有限公司 信息通信分公司,福建 福州 350001;3.北京中电飞华通信有限公司,北京 100070;4.北京创安恒宇科技有限公司,北京 100070
Author(s):
CAI Yuxiang1; 2;  CHEN Lijuan3;  AN Qi4
1.School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China; 2. Information and Communication Branch, State Grid Fujian Electric Power Co., Ltd., Fuzhou 350001, China; 3.Beijing Fibrlink Communications Co., Ltd., Beijing 100070, China; 4.Beijing Chuang’an Hengyu Technology Co., Ltd., Beijing 100070, China
关键词:
缺陷检测;  MobileNetV2;  边缘计算设备;  注意力机制;  SSD
Keywords:
defect detection;  MobileNetV2;  edge computing devices;  attention mechanism;  SSD
分类号:
TP391.4TM7
DOI:
10.13705/j.issn.1671-6833.2025.04.023
文献标志码:
A
摘要:
针对电力物联网中设备表面缺陷自动化检测难题(如破损、污损及人为违规操作导致的缺陷),提出一种面向边缘计算设备的轻量级SSD检测算法。该算法通过3个关键技术创新实现高效检测。首先,在MobileNetV2的瓶颈结构中引入密集连接机制,动态增强图像特征表达能力;其次,基于Non-Local注意力机制构建跨层注意力隐式特征金字塔网络(CL-IFPN),通过与MobileNetV2-SSD的深度融合显著提升小缺陷检测能力;最后,通过在卷积层添加特征融合模块并采用QFL函数,强化不同尺度缺陷的预测精度及正负样本训练平衡性。实验结果表明:在公共数据集VOC2007上,所提算法以79.62%的mAP检测精度和36 帧/s的检测速度表现优于同类算法;在自建电力器件缺陷数据集上,检测性能进一步提升至95.19%的检测精度和24 帧/s的检测速度,充分验证了算法在电力设备缺陷检测场景的实用价值。所提算法为边缘计算环境下的电力物联网设备智能运维提供了有效的技术解决方案。
Abstract:
To address the challenge of automated surface defect detection (e.g., damage, stains, and defects from human violations) in the power IoT, a lightweight SSD detection algorithm for edge computing devices was proposed. The proposed algorithm aimed to achieved efficient detection through three key innovations. Firstly, a dense connection mechanism was introduced into the bottleneck structure of MobileNetV2 to enhance image feature representation dynamically. Secondly, a cross layer attention mechanism implicit feature pyramid network (CL-IFPN) based on No-Local attention mechanism was constructed, and its deep integration with MobileNetV2-SSD significantly improved small-defect detection. Finally, a feature fusion module was added to the convolutional layer, and the QFL function was used to boost prediction accuracy of defects at different sizes and the balance of positive and negative sample training. Experimental results showed that on the public dataset VOC2007, the proposed algorithm achieved a detection accuracy of 79.62% and a speed of 36 frames per second, outperforming similar algorithms. On the self-built power device defect dataset, the detection accuracy reached 95.19% and a speed of 24 frames per second, demonstrating the algorithm′s practicality in power device defect detection. The proposed algorithm offered an effective technical solution for intelligent operation and maintenance of power IoT devices in edge computing environments.

参考文献/References:

[1]QUINLAN J R. Induction of decision trees[J]. Machine Learning, 1986, 1(1): 81-106.

[2]WITTEN I H, FRANK E. HALL M A. Data mining: practical machine learning tools and techniques[M].3rd ed. Amsterdam: Elsevier Inc., 2011.
[3]KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017,60(6): 84-90.
[4]GÜLER R A, NEVEROVA N, KOKKINOS I. DensePose: dense human pose estimation in the wild[EB/OL]. (2018-02-01)[2025-04-02]. https:∥arxiv.org/abs/1802.00434.
[5]GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]∥2014 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2014: 580-587.
[6]CHEN C H, WANG S F, HUANG S Z. An improved faster RCNN-based weld ultrasonic atlas defect detection method[J]. Measurement and Control, 2023, 56(3/4): 832-843.
[7]REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.
[8]REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]∥2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Piscataway: IEEE, 2016: 779-788.
[9]LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[C]∥14th European Conference on Computer Vision, ECCV 2016. Cham: Springer, 2016: 21-37.
[10] SANDLER M, HOWARD A, ZHU M L, et al. MobileNetV2: inverted residuals and linear bottlenecks[C]∥2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 4510-4520.
[11]徐云飞, 张笃周, 王立, 等. 非合作目标局部特征识别轻量化特征融合网络设计[J]. 红外与激光工程, 2020, 49(7): 265-271.
XU Y F, ZHANG D Z, WANG L, et al. Lightweight feature fusion network design for local feature recognition of non-cooperative target[J]. Infrared and Laser Engineering, 2020, 49(7): 265-271.
[12]刘慧, 张礼帅, 沈跃, 等. 基于改进SSD的果园行人实时检测方法[J]. 农业机械学报, 2019, 50(4): 2935, 101.
LIU H, ZHANG L S, SHEN Y, et al. Real-time pedestrian detection in orchard based on improved SSD[J]. Transactions of the Chinese Society for Agricultural Machinery, 2019, 50(4): 29-35, 101.
[13]毛亮, 薛月菊, 朱婷婷, 等. 自然场景下的挖掘机实时监测方法[J]. 农业工程学报, 2020, 36(9): 214-220.
MAO L, XUE Y J, ZHU T T, et al. Method for the realtime monitoring of the excavator in natural scene[J]. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(9): 214-220.
[14]任坤, 黄泷, 范春奇, 等. 基于多尺度像素特征融合的实时小交通标志检测算法[J]. 信号处理, 2020, 36(9): 1457-1463.
REN K, HUANG L, FAN C Q, et al. Real-time small traffic sign detection algorithm based on multi-scale pixel feature fusion[J]. Journal of Signal Processing, 2020, 
[15] LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]∥2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2017: 936-944.
[16] GHIASI G, LIN T Y, LE Q V. NAS-FPN: learning scalable feature pyramid architecture for object detection[C]∥2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2019: 7029-7038.
[17] LI H Y, MIAO S Y, FENG R. DG-FPN: learning dynamic feature fusion based on graph convolution network for object detection[C]∥2020 IEEE International Conference on Multimedia and Expo (ICME). Piscataway: IEEE, 2020: 1-6.
[18] GUO C X, FAN B, ZHANG Q, et al. AugFPN: improving multi-scale feature learning for object detection[C]∥2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2020: 12592-12601.
[19]赵珊, 刘子路, 郑爱玲, 等. 基于MobileNetV2和IFPN改进的SSD垃圾实时分类检测方法[J]. 计算机应用, 2022, 42(增刊1): 106-111.
ZHAO S, LIU Z L, ZHENG A L, et al. Real-time classificaiton and detection method of garbage based on SSD improved with mobileNetV2 and IFPN[J]. Journal of Computer Applications, 2022, 42(S1): 106-111.
[20]种法广, 温蜜, 田英杰, 等. 基于注意力机制的多尺度缺陷绝缘子检测算法[J]. 计算机仿真, 2022, 39(7): 137-142, 147.
CHONG F G, WEN M, TIAN Y J, et al. Multi-scale defect insulator detection algorithm based on attention mechanism[J]. Computer Simulation, 2022, 39(7): 137142, 147.
[21] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]∥2017 IEEE International Conference on Computer Vision (ICCV). Piscataway: IEEE, 2017: 2999-3007.
[22] TONG K, WU Y Q. Rethinking PASCAL-VOC and MSCOCO dataset for small object detection[J]. Journal of Visual Communication and Image Representation, 2023, 93: 103830.

相似文献/References:

[1]廖晓辉,谢子晨,辛忠良,等.基于轻量化YOLOv5的电气设备外部缺陷检测[J].郑州大学学报(工学版),2024,45(04):117.[doi:10.13705/ j.issn.1671-6833.2024.04.010]
　LIAO Xiaohui,XIE Zichen,XIN Zhongliang,et al.Electrical Equipment External Defect Detection Based on Lightweight YOLOv5[J].Journal of Zhengzhou University (Engineering Science),2024,45(01):117.[doi:10.13705/ j.issn.1671-6833.2024.04.010]
[2]韩慧健,邢怀宇,张云峰,等.基于Transformer多元注意力的钢材表面缺陷视觉检测[J].郑州大学学报(工学版),2025,46(05):69.[doi:10.13705/j.issn.1671-6833.2025.05.009]
　HAN Huijian,XING Huaiyu,ZHANG Yunfeng,et al.Visual Detection of Steel Surface Defects Based on Transformer and Multi-attention[J].Journal of Zhengzhou University (Engineering Science),2025,46(01):69.[doi:10.13705/j.issn.1671-6833.2025.05.009]

更新日期/Last Update: 2026-01-17