SavedAdd your journal

Journals / Dicle Üniversitesi Mühendislik Fakültesi Mühendislik Dergisi / 2026 / Cilt: 17 Sayı: 2

Photovoltaic Fault Detection Using SE-MobileNet: A Lightweight Channel Attention Network

Hasan Polat(BİNGÖL ÜNİVERSİTESİ, BİNGÖL TEKNİK BİLİMLER MESLEK YÜKSEKOKULU)0000-0001-5535-4832
Oğuz Kaan Polat(BİNGÖL ÜNİVERSİTESİ)0009-0002-2550-8153
Pages
—
Publication Date
2026-07-28
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

Experimental results demonstrate that this adaptive channel-wise refinement yields a notable improvement in classification performance, increasing the macro F1-score from 0.776 to 0.837, while outperforming a range of benchmark convolutional and transformer-based models.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:The global energy sector is rapidly transitioning from fossil fuels to renewable sources, with solar photovoltaic (PV) systems emerging as a leading solution for sustainable electricity generation.
📊
Key Findings & Data:Therefore, early and accurate fault detection is essential to ensure maintained efficiency and system reliability.
💡
Conclusion & Impact:Experimental results demonstrate that this adaptive channel-wise refinement yields a notable improvement in classification performance, increasing the macro F1-score from 0.776 to 0.837, while outperforming a range of benchmark convolutional and transformer-based models.
🔬Design:Modeling & Simulation
📍Scope:Comprehensive Sample
⚙️Variables:Renewable energy · Solar energy · Photovoltaic

Abstract

The global energy sector is rapidly transitioning from fossil fuels to renewable sources, with solar photovoltaic (PV) systems emerging as a leading solution for sustainable electricity generation. However, maintaining the cleanliness and structural integrity of PV panels is crucial for optimal operation. Therefore, early and accurate fault detection is essential to ensure maintained efficiency and system reliability. This study introduces a lightweight deep learning framework, termed SE-MobileNet, for diagnosing PV surface faults such as dust and bird droppings, as well as physical and electrical damage. The proposed model incorporates a Squeeze-and-Excitation (SE) block into a MobileNet backbone to enable channel-wise feature recalibration without imposing a significant computational burden. Experimental results demonstrate that this adaptive channel-wise refinement yields a notable improvement in classification performance, increasing the macro F1-score from 0.776 to 0.837, while outperforming a range of benchmark convolutional and transformer-based models. Furthermore, the model achieves a favorable trade-off between accuracy and computational complexity, with an inference time of 12.73 ms and 78.54 FPS, enabling practical deployment under near real-time conditions in PV monitoring systems.

Özet

The global energy sector is rapidly transitioning from fossil fuels to renewable sources, with solar photovoltaic (PV) systems emerging as a leading solution for sustainable electricity generation. However, maintaining the cleanliness and structural integrity of PV panels is crucial for optimal operation. Therefore, early and accurate fault detection is essential to ensure maintained efficiency and system reliability. This study introduces a lightweight deep learning framework, termed SE-MobileNet, for diagnosing PV surface faults such as dust and bird droppings, as well as physical and electrical damage. The proposed model incorporates a Squeeze-and-Excitation (SE) block into a MobileNet backbone to enable channel-wise feature recalibration without imposing a significant computational burden. Experimental results demonstrate that this adaptive channel-wise refinement yields a notable improvement in classification performance, increasing the macro F1-score from 0.776 to 0.837, while outperforming a range of benchmark convolutional and transformer-based models. Furthermore, the model achieves a favorable trade-off between accuracy and computational complexity, with an inference time of 12.73 ms and 78.54 FPS, enabling practical deployment under near real-time conditions in PV monitoring systems.

🔖

Cite This Article

Polat, H., Polat, O. K. (2026). Photovoltaic Fault Detection Using SE-MobileNet: A Lightweight Channel Attention Network. Dicle Üniversitesi Mühendislik Fakültesi Mühendislik Dergisi, Cilt: 17 Sayı: 2. https://doi.org/10.24012/dumf.1926748

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (32)

  1. S. Shaik et al., “Experimental analysis on the impacts of soil deposition and bird droppings on the thermal performance of photovoltaic panels,” Case Stud. Therm. Eng., vol. 48, no. March, 2023, doi: 10.1016/j.csite.2023.103128.
  2. R. A. M. Rudro et al., “SPF-Net: Solar panel fault detection using U-Net based deep learning image classification,” Energy Reports, vol. 12, no. July, pp. 1580–1594, 2024, doi: 10.1016/j.egyr.2024.07.044.
  3. A. Sohail, N. Ul Islam, A. Ul Haq, S. Ul Islam, I. Shafi, and J. Park, “Fault detection and computation of power in PV cells under faulty conditions using deep-learning,” Energy Reports, vol. 9, pp. 4325–4336, 2023, doi: 10.1016/j.egyr.2023.03.094.
  4. A. Rodrigo, I. Munasinghe, and A. Perera, “Vision-Based Object Detection for UAV Solar Panel Inspection Using an Enhanced Defects Dataset,” 2025, [Online]. Available: http://arxiv.org/abs/2509.05348
  5. M. Jalal, I. U. Khalil, and A. ul Haq, “Deep learning approaches for visual faults diagnosis of photovoltaic systems: State-of-the-Art review,” Results Eng., vol. 23, no. July, p. 102622, 2024, doi: 10.1016/j.rineng.2024.102622.
  6. Y. Y. Hong and R. A. Pula, “Methods of photovoltaic fault detection and classification: A review,” Energy Reports, vol. 8, pp. 5898–5929, 2022, doi: 10.1016/j.egyr.2022.04.043.
  7. K. Masita, A. Hasan, T. Shongwe, and H. A. Hilal, “Deep learning in defects detection of PV modules: A review,” Sol. Energy Adv., vol. 5, no. September 2024, p. 100090, 2025, doi: 10.1016/j.seja.2025.100090.
  8. M. N. Abuhashish, A. Refaat, A. Kalas, M. S. Hamad, and M. H. Elfar, “Towards accurate and reliable fault diagnosis in PV systems: Techniques, challenges, and future directions,” Process Saf. Environ. Prot., vol. 198, no. February, p. 107217, 2025, doi: 10.1016/j.psep.2025.107217.
  9. J. Hu, L. Shen, S. Albanie, G. Sun, and E. Wu, “Squeeze-and-Excitation Networks,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 42, no. 8, pp. 2011–2023, 2020, doi: 10.1109/TPAMI.2019.2913372.
  10. A. G. Howard et al., “MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications,” Apr. 2017, [Online]. Available: http://arxiv.org/abs/1704.04861
  11. R. Cavieres, R. Barraza, D. Estay, J. Bilbao, and P. Valdivia-Lefort, “Automatic soiling and partial shading assessment on PV modules through RGB images analysis,” Appl. Energy, vol. 306, no. October 2021, 2022, doi: 10.1016/j.apenergy.2021.117964.
  12. K. Niazi, W. Akhtar, H. A. Khan, S. Sohaib, and A. K. Nasir, “Binary Classification of Defective Solar PV Modules Using Thermography,” 2018 IEEE 7th World Conf. Photovolt. Energy Conversion, WCPEC 2018 - A Jt. Conf. 45th IEEE PVSC, 28th PVSEC 34th EU PVSEC, no. 1, pp. 753–757, 2018, doi: 10.1109/PVSC.2018.8548138.
  13. A. Et-taleby, Y. Chaibi, N. Ayadi, B. Elkari, M. Benslimane, and Z. Chalh, “Enhancing fault detection and classification in photovoltaic systems based on a hybrid approach using fuzzy logic algorithm and thermal image processing,” Sci. African, vol. 28, no. November 2024, p. e02684, 2025, doi: 10.1016/j.sciaf.2025.e02684.
  14. W. M. Shaban, “Detection and classification of photovoltaic module defects based on artificial intelligence,” Neural Comput. Appl., vol. 36, no. 27, pp. 16769–16796, 2024, doi: 10.1007/s00521-024-10000-z.
  15. M. U. Ali, H. F. Khan, M. Masud, K. D. Kallu, and A. Zafar, “A machine learning framework to identify the hotspot in photovoltaic module using infrared thermography,” Sol. Energy, vol. 208, no. May, pp. 643–651, 2020, doi: 10.1016/j.solener.2020.08.027.
  16. D. Marletta, A. Midolo, and E. Tramontana, “Detecting Photovoltaic Panels in Aerial Images by Means of Characterising Colours,” Technologies, vol. 11, no. 6, 2023, doi: 10.3390/technologies11060174.
  17. I. Ibn Malek and H. Imtiaz, “Advanced fault detection in PV panels using deep neural networks: leveraging transfer learning and electroluminescence image processing,” Energy Adv., pp. 180–193, 2025, doi: 10.1039/d5ya00239g.
  18. P. R. Sonawane, S. Bhandari, R. B. Patil, and S. Al-Dahidi, “Comparative analysis of methods used to solve class imbalance problem in dataset for CNN-based fault diagnosis of solar photovoltaic panels,” Int. J. Thermofluids, vol. 28, no. June 2025, p. 101303, 2025, doi: 10.1016/j.ijft.2025.101303.
  19. R. Tang, Z. Ren, S. Ning, and Y. Zhang, “Fault classification of photovoltaic module infrared images based on transfer learning and interpretable convolutional neural network,” Sol. Energy, vol. 276, no. March, p. 112703, 2024, doi: 10.1016/j.solener.2024.112703.
  20. O. F. Bamisile et al., “Attention residual network with multi-scale convolution branch for efficient solar photovoltaic module defect classification,” Sol. Energy, vol. 307, no. December 2025, p. 114323, 2026, doi: 10.1016/j.solener.2026.114323.
  21. Y. Ledmaoui, A. El Maghraoui, M. El Aroussi, and R. Saadane, “Enhanced Fault Detection in Photovoltaic Panels Using CNN-Based Classification with PyQt5 Implementation,” Sensors, vol. 24, no. 22, pp. 1–20, 2024, doi: 10.3390/s24227407.
  22. “PV Panel Defect Dataset.” Accessed: Apr. 07, 2026. [Online]. Available: https://www.kaggle.com/datasets/alicjalena/pv-panel-defect-dataset
  23. K. Awedat, G. Comert, M. Ayad, and A. Mrebit, “Advanced fault detection in photovoltaic panels using enhanced U-Net architectures,” Mach. Learn. with Appl., vol. 20, no. February, p. 100636, 2025, doi: 10.1016/j.mlwa.2025.100636.
  24. H. Polat, “Time-Frequency Complexity Maps for EEG-Based Diagnosis of Alzheimer ’ s Disease Using a Lightweight Deep Neural Network Traitement du Signal Time-Frequency Complexity Maps for EEG- Based Diagnosis of Alzheimer ’ s Disease Using a Lightweight Deep Neural N,” no. December 2022, 2023, doi: 10.18280/ts.390623.
  25. H. Polat, “DAMMD-Net: A lightweight and enhanced deep segmentation network for skin lesion detection,” Digit. Signal Process. A Rev. J., vol. 167, no. 1, p. 105477, 2025, doi: 10.1016/j.dsp.2025.105477.
  26. G. Huang, Z. Liu, L. Van Der Maaten, and K. Q. Weinberger, “Densely connected convolutional networks,” Proc. - 30th IEEE Conf. Comput. Vis. Pattern Recognition, CVPR 2017, vol. 2017-Janua, pp. 2261–2269, 2017, doi: 10.1109/CVPR.2017.243.
  27. K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit., vol. 2016-Decem, pp. 770–778, 2016, doi: 10.1109/CVPR.2016.90.
  28. M. Tan and Q. V. Le, “EfficientNet: Rethinking model scaling for convolutional neural networks,” 36th Int. Conf. Mach. Learn. ICML 2019, vol. 2019-June, pp. 10691–10700, 2019.
  29. A. Dosovitskiy et al., “an Image Is Worth 16X16 Words: Transformers for Image Recognition At Scale,” ICLR 2021 - 9th Int. Conf. Learn. Represent., 2021.
  30. A. Khalifeh and R. Ghanem, “Solar Panel Soiling Classification Using Machine Learning Algorithms,” 2024 25th Int. Arab Conf. Inf. Technol. ACIT 2024, pp. 1–5, 2024, doi: 10.1109/ACIT62805.2024.10877104.
  31. M. R. Shoaib, H. M. Emara, J. Zhao, M. T. Ahvanooey, and E. Nabil, “Advanced deep learning approaches for fault detection in solar PV systems: A comparative study of SPDA and AIFD-SolDL,” Appl. Soft Comput., vol. 184, no. PB, p. 113592, 2025, doi: 10.1016/j.asoc.2025.113592.
  32. M. Eshwaran Ramachandran, G. Babu Balachandran, P. Velladurai, and A. Rajakumar, “Deep learning for solar PV fault classification using RGB imaging and comparison of preprocessing techniques,” Sol. Energy, vol. 301, no. May, p. 113959, 2025, doi: 10.1016/j.solener.2025.113959.