PHOTOVOLTAIC MODULE DEFECT DETECTION METHOD BASED ON IMPROVED YOLOV5S NETWORK

Ren Xiwei; Yu Jie; Han Xin; Li Zhaoyun; Yang Menglu; He Lifeng

doi:10.19912/j.0254-0096.tynxb.2023-1934

PDF(3700 KB)

Acta Energiae Solaris Sinica ›› 2025, Vol. 46 ›› Issue (3) : 428-434. DOI: 10.19912/j.0254-0096.tynxb.2023-1934

PHOTOVOLTAIC MODULE DEFECT DETECTION METHOD BASED ON IMPROVED YOLOV5S NETWORK

Ren Xiwei^1,2, Yu Jie^1,2, Han Xin¹, Li Zhaoyun¹, Yang Menglu¹, He Lifeng¹

Author information +

History +

Abstract

Given the dense arrangement and small target size of photovoltaic modules, which make defect detection and maintenance challenging and directly impact the power generation efficiency of photovoltaic modules, an improved method for detecting defects in photovoltaic modules using the YOLOv5s network is proposed. Firstly, considering the size charact eristics of defects in infrared images of photovoltaic modules, the K-means++ algorithm is used to re-cluster defect targets, determining appropriate anchor box sizes to ensure that the clustered anchor boxes better fit the characteristics of small targets. Then, a simplified BiFPN is employed to fuse more features, with a multi-path residual connection module added before fusion to enhance sensitivity to small target photovoltaic defects. Furthermore, the YOLOv5s backbone network is merged, simplifying the network structure, reducing the number of downsampling operations, increasing image resolution, and enriching information on small target features. Subsequently, the S_IoU loss function is introduced into the YOLOv5s architecture to improve network performance, enabling deployment on small, lightweight devices. Finally, the improved YOLOv5s network is tested on a self-built dataset of infrared images of photovoltaic module defects. Experimental results demonstrate that the improved YOLOv5s network for photovoltaic module defect detection outperforms the baseline methods. Compared to the original network, the improved network’s m_AP@0.5 increased by 1.7%, reaching an FPS of 46.3, fully meeting the practical requirements of photovoltaic module defect detection.

Key words

defect detection / YOLOv5 / loss function / small object enhancement / photovoltaic module / infrared images

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

Ren Xiwei, Yu Jie, Han Xin, Li Zhaoyun, Yang Menglu, He Lifeng. PHOTOVOLTAIC MODULE DEFECT DETECTION METHOD BASED ON IMPROVED YOLOV5S NETWORK[J]. Acta Energiae Solaris Sinica. 2025, 46(3): 428-434 https://doi.org/10.19912/j.0254-0096.tynxb.2023-1934

References

[1] 姚俊, 陈彦涛, 阮晓娜, 等. “双碳” 目标下我国光伏发电的装机容量预测[J]. 节能, 2023, 42(10): 36-38.
YAO J, CHEN Y T, RUAN X N, et al.Prediction of installed capacity of Chinese photovoltaic power generation market under the target of ‘double carbon’[J]. Energy conservation, 2023, 42(10): 36-38.
[2] 魏晨晨, 王超, 李想, 等. 光伏组件发电效率影响因素分析[J]. 建设科技, 2023(1): 79-81.
WEI C C, WANG C, LI X, et al.Analysis of influencing factors on power generation efficiency of photovoltaic modules[J]. Construction science and technology, 2023(1): 79-81.
[3] SONG X Y, YANG L, XIA X H, et al.Defect detection method for solar cells based on human visual characteristics[C]//2020 5th International Conference on Mechanical, Control and Computer Engineering (ICMCCE). Harbin, China, 2020: 515-518.
[4] 邓枭, 叶蔚, 谢睿, 等. 基于深度学习的源代码缺陷检测研究综述[J]. 软件学报, 2023, 34(2): 625-654.
DENG X, YE W, XIE R, et al.Survey of source code bug detection based on deep learning[J]. Journal of software, 2023, 34(2): 625-654.
[5] 周颖, 叶红, 王彤, 等. 基于多尺度CNN的光伏组件缺陷识别[J]. 太阳能学报, 2022, 43(2): 211-216.
ZHOU Y, YE H, WANG T, et al.Photovoltaic module defect identification based on mulit-scale convolution neural network[J]. Acta energiae solaris sinica, 2022, 43(2): 211-216.
[6] ALI M U, KHAN H F, MASUD M, et al.A machine learning framework to identify the hotspot in photovoltaic module using infrared thermography[J]. Solar energy, 2020, 208: 643-651.
[7] 钟泳松, 徐凌桦, 周克. 基于改进SSD算法的光伏组件缺陷检测研究[J]. 微处理机, 2022, 43(1): 22-25.
ZHONG Y S, XU L H, ZHOU K.Research on defect detection of photovoltaic module based on improved SSD algorithm[J]. Microprocessors, 2022, 43(1): 22-25.
[8] 王淑青, 张鹏飞, 要若天, 等. 基于改进YOLOv5的太阳能电池片表面缺陷检测[J]. 仪表技术与传感器, 2022(5): 111-116.
WANG S Q, ZHANG P F, YAO R T, et al.Surface defect detection of solar cell based on improved YOLOv5[J]. Instrument technique and sensor, 2022(5): 111-116.
[9] 郭岚, 刘正新. 基于改进YOLOv5的光伏组件缺陷检测[J]. 激光与光电子学进展, 2023, 60(20): 148-156.
GUO L, LIU Z X.Improved YOLOv5-based defect detection in photovoltaic modules[J]. Laser & optoelectronics progress, 2023, 60(20): 148-156.
[10] TAN M X, PANG R M, LE Q V.EfficientDet: scalable and efficient object detection[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, WA, USA, 2020: 10778-10787.
[11] 杭伯安. 基于改进K-means算法的沥青路面裂缝分割方法[J]. 公路交通科技, 2023, 40(4): 1-8.
HANG B A.A method for asphalt pavement crack segmentation based on improved K-means algorithm[J]. Journal of highway and transportation research and development, 2023, 40(4): 1-8.
[12] WANG K X, LIEW J H, ZOU Y T, et al.PANet: few-shot image semantic segmentation with prototype alignment[C]//2019 IEEE/CVF International Conference on Computer Vision (ICCV). Seoul, Korea (South), 2019: 9196-9205.
[13] SRIVASTAVA A, CHANDA S, JHA D, et al.PAANet: progressive alternating attention for automatic medical image segmentation[C]//2021 4th International Conference on Bio-Engineering for Smart Technologies (BioSMART). Paris, France, 2021: 1-4.
[14] 王银, 高瑞泽, 李茂环, 等. 基于YOLOv5 LiteX算法的光伏组件缺陷检测[J]. 太阳能学报, 2023, 44(9): 101-108.
WANG Y, GAO R Z, LI M H, et al.Photovoltaic modules defect detection based on YOLOv5 LiteX algorithm[J]. Acta energiae solaris sinica, 2023, 44(9): 101-108.
[15] 蒋兴群, 刘波, 宋力, 等. 基于改进YOLO-v3的风力机叶片表面损伤检测识别[J]. 太阳能学报, 2023, 44(3): 212-217.
JIANG X Q, LIU B, SONG L, et al.Surface damage detection and recognition of wind turbine blade based on improved YOLO-v3[J]. Acta energiae solaris sinica, 2023, 44(3): 212-217.
[16] 孔松涛, 徐甄泽, 林星宇, 等. 基于改进YOLOv5算法的光伏组件红外热成像缺陷检测[J]. 红外技术, 2023, 45(9): 974-981.
KONG S T, XU Z Z, LIN X Y, et al.Infrared thermal imaging defect detection of photovoltaic module based on improved YOLOv5 algorithm[J]. Infrared technology, 2023, 45(9): 974-981.
[17] GEVORGYAN Z. SIoU loss: more powerful learning for bounding box regression[J]. arXiv preprint arXiv:2205. 12740, 2022.
[18] 刘雄彪, 杨贤昭, 陈洋, 等. 基于CIoU改进边界框损失函数的目标检测方法[J]. 液晶与显示, 2023, 38(5): 656-665.
LIU X B, YANG X Z, CHEN Y, et al.Object detection method based on CIoU improved bounding box loss function[J]. Chinese journal of liquid crystals and displays, 2023, 38(5): 656-665.
[19] 麻斯亮, 许勇. 最小点距离的边界框回归损失函数及其应用[J]. 小型微型计算机系统, 2024, 45(11):2695-2701.
MA S L, XU Y.Bounding box regression loss function based on minimum point distance and its application[J]. Journal of Chinese computer systems, 2024, 45(11): 2695-2701.
[20] 叶翔, 孙嘉兴, 甘永叶, 等. 改进YOLOv3模型在无人机巡检输电线路部件缺陷检测中的应用研究[J]. 电测与仪表, 2023, 60(5): 85-91.
YE X, SUN J X, GAN Y Y, et al.Application of improved YOLOv3 model in defect detection of transmission line components in UAV patrol inspection[J]. Electrical measurement & instrumentation, 2023, 60(5): 85-91.
[21] 任喜伟, 韩欣, 钟弋, 等. 基于改进U-Net网络的光伏板图像分割方法[J]. 陕西科技大学学报, 2023, 41(2): 155-161.
REN X W, HAN X, ZHONG Y, et al.Photovoltaic panel image segmentation method based on improved U-Net network[J]. Journal of Shaanxi University of Science & Technology, 2023, 41(2): 155-161.
[22] LIU W, ANGUELOV D, ERHAN D, et al.SSD: single shot MultiBox detector[M].Lecture Notes in Computer Science. Cham: Springer International Publishing, 2016: 21-37.
[23] REDMON J, FARHADI A. YOLOv3: an incremental improvement[J]. ArXiv preprint arXiv:1804.02767, 2018.