Design of a Photovoltaic Panel Defect Detection System Based on the YOLO Algorithm_Vol. 4 No. 2 (JBDC 2026)_Journal of Big Data and Computing (ISSN: 2959-0590)

Home > Journal of Big Data and Computing (ISSN: 2959-0590) > Vol. 4 No. 2 (JBDC 2026) >

Design of a Photovoltaic Panel Defect Detection System Based on the YOLO Algorithm

DOI: https://doi.org/10.62517/jbdc.202601220

Author(s)

Jingjing Liu, Yihuan Zhang, Zikang Shao, Xinyu Zhang, Jingyi Yang, Dongyang Zhang, Boyang Zhang, Ruian Yan*

Affiliation(s)

School of Artificial Intelligence and Big Data, Henan University of Technology, Zhengzhou, Henan, China *Corresponding Author

Abstract

With the rapid expansion of the scale of photovoltaic power plants, the industry operation and maintenance links put forward higher requirements for the efficiency, real-time and intelligent level of defect detection. However, traditional inspection methods mostly rely on manual observation or simple monitoring methods, which have problems such as low detection efficiency, limited real-time analysis capabilities, and difficulty in structured management of defect information. To this end, this paper designs and implements an intelligent monitoring system for photovoltaic modules, which integrates defect detection, target tracking, multi-modal analysis and digital twin visual management functions to improve the efficiency and reliability of power station inspection and operation and maintenance decision-making. The system uses a front-end and back-end separation architecture: the front-end builds an interactive visual interface based on the Vue 3 framework and TypeScript language technology; the back-end uses the Flask framework to build RESTful services, which are responsible for data processing and business collaboration. In order to improve the security and engineering availability of the system, the system introduces the JWT authentication mechanism, and adopts the asynchronous task scheduling mechanism based on Celery to support time-consuming reasoning and batch processing tasks. At the same time, the whole system enhances maintainability and scalability through modular design. In terms of defect analysis, the system integrates YOLO series target detection algorithm model, ByteTrack tracking algorithm and Qwen-VL-Plus multi-modal model to realize the recognition, location and continuous tracking of targets such as bird droppings, cracks, stains and panel areas. The experimental results show that in the crack detection task, the model achieves a performance of mAP @ 0.5 of 0.773 and a recall rate of 0.827. In order to adapt to the edge deployment scenario, the system further combines optimization strategies such as TensorRT acceleration, ONNX Runtime reasoning, model quantization and image enhancement to significantly improve reasoning efficiency and operational stability. The system can provide effective technical support for intelligent monitoring and early fault warning of photovoltaic power plants.

Keywords

Photovoltaic Module Defect Detection; Intelligent Monitoring System; YOLO; ByteTrack; Multimodal Analysis; Edge Deployment

References

[1]Vue.js Team, Vue. js v3.5 Documentation, 2024. [Online]. Available: https:// vuejs.org/. [2]Microsoft, TypeScript 5.5 Handbook, 2024. [Online]. Available: https:// www. typescriptlang.org/. [3]Element Plus Team, Element Plus 2.x Documentation, 2024. [Online]. Available: https:// element-plus.org/. [4]E. You, Pinia: The intuitive store for Vue.js, 2024. [Online]. Available: https:// pinia.vuejs.org/. [5]E. You, Vue Router 4, 2024. [Online]. Available: https:// router.vuejs.org/. [6]Axios Contributors, Axios HTTP Client, 2024. [Online]. Available: https:// axios-http.com/. [7]Apache ECharts Team, ECharts 5.5 Documentation, 2024. [Online]. Available: https:// echarts.apache.org/. [8]X. Sun et al., "A Web-Based Visualization System for Smart Factory Monitoring Using ECharts,"2024 IEEE 7th International Conference on Computer and Communication Systems (ICCCS), 2024, pp. 123-128. [9]Bilibili Inc., flv. js: HTML5 Flash Video (FLV) Player, GitHub repository, 2024. [Online]. Available:https:// github.com/ bilibili/ flv. js. [10]M. Deitsch, C. Buerhop-Lutz, A. Maier, F. Gallwitz, and C. Riess, "Automatic classification of defective photovoltaic module cells in electroluminescence images," Solar Energy, vol. 185, pp. 455–468, 2019, doi: 10.1016/ j.solener.2019.02.067. [11]J. Redmon and A. Farhadi,"YOLOv3: An incremental improvement," arXiv preprint arXiv: 1804.02767, 2018, doi: 10.48550/arXiv.1804.02767. [12]Y. Zhang, P. Sun, Y. Jiang, D. Yu, Z. Yuan, P. Luo, W. Liu, and X. Wang, "ByteTrack: Multi-object tracking by associating every detection box," in Proc. European Conf. Computer Vision (ECCV), Cham, Switzerland: Springer, 2022, pp. 1–21, doi: 10.1007/978-3-031-20047-2_17. [13]N. Dragoni, S. Giallorenzo, A. Lluch-Lafuente, M. Mazzara, F. Montesi, R. Mustafin, and L. Safina," Microservices: Yesterday, today, and tomorrow," in Present and Ulterior Software Engineering, Cham, Switzerland: Springer, 2017, pp. 195–216, doi: 10.1007/978-3-319-67425-4_12. [14]D. R. Hipp, "SQLite," Proceedings of the VLDB Endowment, vol. 3, no. 1–2, pp. 1723–1724, 2010, doi: 10.14778/1920841.1920887. [15]Redmon J, Divvala S, Girshick R, et al. You Only Look Once: Unified, Real-Time Object Detection//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, NV, USA: IEEE, 2016: 779-788. [16]Ren S, He K, Girshick R, et al. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks//Advances in Neural Information Processing Systems. 2015, 28: 91-99. [17]Bochkovskiy A, Wang C Y, Liao H Y M. YOLOv4: Optimal Speed and Accuracy of Object Detection. arXiv: 2004.10934, 2020. [18]Zhang, T., Cui, B. W., & Cui, X. M. (2026). Photovoltaic panel defect detection algorithm based on improved YOLOv11n. Electronic Measurement Technology, 1–12. Retrieved from https:// link.cnki.net/urlid/11.2175.TN. 20260130.1931.012. [19]Zhang H, Cisse M, Dauphin Y N, Lopez-Paz D. mixup: Beyond Empirical Risk Minimization// International Conference on Learning Representations (ICLR). 2018. [20]Lin T Y, Goyal P, Girshick R, He K, Dollár P. Focal Loss for Dense Object Detection/ /Proceedings of the IEEE International Conference on Computer Vision (ICCV). 2017: 2999-3007. [21]Pan S J, Yang Q. A Survey on Transfer Learning. IEEE Transactions on Knowledge and Data Engineering, 2010, 22(10): 1345-1359. [22]Zheng Z, Wang P, Liu W, Li J, Ye R, Ren D. Distance-IoU Loss: Faster and Better Learning for Bounding Box Regression// Proceedings of the AAAI Conference on Artificial Intelligence. 2020, 34(07): 12993-13000. [23]Khanam R, Asghar T, Hussain M. Comparative Performance Evaluation of YOLOv5, YOLOv8, and YOLOv11 for Solar Panel Defect Detection. Solar, 2025, 5(1): 6. [24]Zhang Y, Sun P, Jiang Y, et al. ByteTrack: Multi-Object Tracking by Associating Every Detection Box//European Conference on Computer Vision. Cham: Springer, 2022: 1-21.