Yeni Bir CNN-PCA-SVM Derin Hibrit Modeline Dayalı Arızalı Fotovoltaik Modül Hücrelerinin Otomatik Sınıflandırılması

Abstract

Günümüzde sürdürülebilir enerji üretimi için fotovoltaik (PV) enerji santrallerinin hızlı gelişimine olanak sağlamıştır. Enerji üretiminin sürekliliğinin sağlanması için bakım ve arıza tespiti önemli bir rol oynamaktadır. PV modüllerinin elektrolüminesans (EL) görüntülerinin manuel olarak incelenmesi, büyük bir insan gücü ve zaman yatırımını gerektirir. Bu çalışmada, EL görüntülerde PV hücrelerinin otomatik arıza tespiti için hibrit derin özniteliklerin, temel bileşenler analizi (PCA) öznitelik seçme algoritması ile optimize edilen bir yöntem sunmaktadır. Evrişimsel sinir ağı (CNN) mimarilerinin güçlü yönlerini birleştiren, hafif ve yüksek performanslı bir model önerilmektedir. İlk olarak EL görüntülerini içeren veri setindeki arızalı ve işlevsel sınıflarına ait veri dengesizliğinden dolayı veri arttırma teknikleri kullanılmıştır. MobileNetV2, DenseNet201 ve InceptionV3 CNN modellerine entegre edilen PCA algoritması ile hibrit kullanılarak gerçekleştirilen deneysel çalışmalarda doğruluk, kesinlik, duyarlılık ve F1-skoru değerleri sırasıyla %92,19, %92, %90 ve %91 olarak elde edilmiştir. Sonuçlar analiz edildiğinde, önerilen yöntemin PV panel hücrelerindeki arızaların tespitinde etkili bir performansa sahip olduğu gözlemlenmiştir.

Keywords

• Derin öğrenme
• fotovoltaik hücreler
• kusur tespiti
• özellik seçimi

Automatic Classification of Defective Photovoltaic Module Cells Based on a Novel CNN-PCA-SVM Deep Hybrid Model in Electroluminescence Images

Abstract

In today’s world, the rapid development of photovoltaic (PV) power plants has facilitated sustainable energy production. Maintenance and defect detection play crucial roles in ensuring the continuity of energy production. The manual inspection of electroluminescence (EL) images of PV modules requires significant human power and time investment. This study presents a method for the automatic fault detection of PV cells in EL images using hybrid deep features optimized with a principal component analysis (PCA) feature selection algorithm. A lightweight and high-performance model that combines the strengths of convolutional neural network (CNN) architectures was proposed. First, data augmentation techniques were employed owing to the imbalance between the defective and functional classes in the dataset containing EL images. In experimental studies conducted by integrating the PCA algorithm into MobileNetV2, DenseNet201, and InceptionV3 CNN models, accuracy, precision, recall, and F1-score values of 92.19%, 92%, 90%, and 91%, respectively, were achieved. When the results were analyzed, it was observed that the proposed method was effective in detecting faults in PV panel cells.

Keywords

• Deep learning
• photovoltaic cells
• defect detection
• feature selection

References

1. Pillai DS, and Rajasekar N. A comprehensive review on protection challenges and fault diagnosis in PV systems. Renewable and Sustainable Energy Reviews, 2018; 91:18–40.
2. Eltamaly AM. A novel benchmark shading pattern for PV maximum power point trackers evaluation. Sol Energy, 2023; 263:111897.
3. Demir A, Dinçer AE, and Yılmaz K. A novel method for the site selection of large-scale PV farms by using AHP and GIS: A case study in İzmir, Türkiye. Sol Energy, 2023;259:235–245.
4. Omazic A et al. Relation between degradation of polymeric components in crystalline silicon PV module and climatic conditions: A literature review. Sol Energy Mater Sol Cells, 2019;192:123–133.
5. Vázquez M, and Rey-Stolle I. Photovoltaic module reliability model based on field degradation studies. Prog Photovoltaics Res Appl, 2008;16(5):419–433.
6. Silvestre S, Kichou S, Chouder A, Nofuentes G, and Karatepe E. Analysis of current and voltage indicators in grid connected PV (photovoltaic) systems working in faulty and partial shading conditions. Energy, 2015;86:42–50.
7. Dhoke A, Sharma R, and Saha TK. PV module degradation analysis and impact on settings of overcurrent protection devices. Sol Energy, 2018;160:360–367.
8. Wang H, Zhao J, Sun Q, and Zhu H. Probability modeling for PV array output interval and its application in fault diagnosis. Energy, 2019;189:116248.

9. Gong B, An A, Shi Y, and Zhang X. Fast fault detection method for photovoltaic arrays with adaptive deep multiscale feature enhancement. Appl Energy, 2024;353:122071.
10. Munoz MA, Alonso-García MC, Vela N and Chenlo F. Early degradation of silicon PV modules and guaranty conditions. Sol Energy, 2011;85(9):264–2274.
11. Djordjevic S, Parlevliet D, and Jennings P. Detectable faults on recently installed solar modules in Western Australia. Renew Energy,2014; 67:215–221
12. Dhoke A, Sharma R, and Saha TK. PV module degradation analysis and impact on settings of overcurrent protection devices. Sol. Energy, 2018;160:60–367.
13. Dhimish M. Micro cracks distribution and power degradation of polycrystalline solar cells wafer: Observations constructed from the analysis of 4000 samples. Renew Energy, 2020;145:466–477.
14. Abdelhamid M, Singh R, and Omar M. Review of microcrack detection techniques for silicon solar cells. IEEE J Photovolt, 2014;4(1):514–524.
15. Tsanakas JA, Ha L, and Buerhop C. Faults and infrared thermographic diagnosis in operating c-Si photovoltaic modules: A review of research and future challenges. Renewable Sustainable Energy Rev., 2016;62:695–709.
16. Fuyuki T, and Kitiyanan A. Photographic diagnosis of crystalline silicon solar cells utilizing electroluminescence. Appl Phys A Mater Sci Process,2009;96(1):189–196.
17. Breitenstein O et al. Can luminescence imaging replace lock-in thermography on solar cells. IEEE J Photovolt, 2011;1(2):159–167.
18. Tsanakas JA, Chrysostomou D, Botsaris PN, and Gasteratos A. Fault diagnosis of photovoltaic modules through image processing and Canny edge detection on field thermographic measurements. Int J Sustainable Energy, 2013;34(6):351–372.
19. Pratt L, Govender D, and Klein R. Defect detection and quantification in electroluminescence images of solar PV modules using U-net semantic segmentation. Renew Energy, 2021;178:1211–1222.
20. Vidal De Oliveira AK, Rüther R, and Aghaei M. Automatic Fault Detection of Photovoltaic Arrays by Convolutional Neural Networks During Aerial Infrared Thermography. In Proceedings of the 36th European Photovoltaic Solar Energy Conference and Exhibition,2019;9-13.
21. Deitsch S et al. Automatic classification of defective photovoltaic module cells in electroluminescence images. Sol Energy, 2019;185:455–468.
22. Tang W, Yang Q, Xiong K, and Yan W. Deep learning based automatic defect identification of photovoltaic module using electroluminescence images. Sol Energy, 2020;201:453–460.
23. Hong F, Song J, Meng H, Rui W, Fang F, and Guangming Z. A novel framework on intelligent detection for module defects of PV plant combining the visible and infrared images. Sol Energy, 2022;236:406–416.
24. Zhao Y, Zhan K, Wang Z, and Shen W. Deep learning-based automatic detection of multitype defects in photovoltaic modules and application in real production line. Prog Photovoltaics Res Appl., 2021;29(4):471–484.
25. Moradi Sizkouhi A, Aghaei M, and Esmailifar SM. A deep convolutional encoder-decoder architecture for autonomous fault detection of PV plants using multi-copters. Sol Energy, 2021;223:217–228.
26. Sun T, Xing H, Cao S, Zhang Y, Fan S, and Liu P. A novel detection method for hot spots of photovoltaic (PV) panels using improved anchors and prediction heads of YOLOv5 network. Energy Reports, 2022;8:1219-1229.
27. Yanilmaz S, Türkoğlu M, and Aslan M. Güneş enerjisi santrallerinde YOLO algoritmaları ile hotspot kusurlarının tespiti. Fırat University Journal of Engineering Science, 2024;36(1):121-132.
28. Cao Y, et al. Improved YOLOv8-GD deep learning model for defect detection in electroluminescence images of solar photovoltaic modules, Eng Appl Artif Intell, 2024;131:107866.
29. Açikgöz H, and Korkmaz D. Elektrolüminesans görüntülerde arızalı fotovoltaik panel hücrelerin evrişimli sinir ağı ile otomatik sınıflandırılması. Fırat University Journal of Engineering Science, 2022;34(2):589–600.
30. Demirci MY, Beşli N, and Gümüşçü A. Defective PV cell detection using deep transfer learning and EL imaging. In International Conference on Data Science, Machine Learning and Statistics 2019 (DMS-2019), 2019;311–314.
31. Demirci MY, Beşli N, and Gümüşçü A. Efficient deep feature extraction and classification for identifying defective photovoltaic module cells in Electroluminescence images, Expert Syst Appl, 2021;175:114810.
32. Deitsch S et al. Segmentation of photovoltaic module cells in uncalibrated electroluminescence images. Mach Vis Appl, 2021;32(4):1–23.
33. Imak A, Celebi A, Siddique K, Turkoglu M, Sengur A, and Salam I. Dental caries detection using score-based multi-input deep convolutional neural network. IEEE Access, 2022;10:18320–18329.
34. Turkoglu M. COVIDetectioNet: COVID-19 diagnosis system based on X-ray images using features selected from pre-learned deep features ensemble. Appl Intell, 2021;51(3)1213–1226.
35. Szegedy C, Vanhoucke V, Ioffe S, Shlens J, and Wojna Z. Rethinking the inception architecture for computer vision. In Proceedings of the IEEE conference on computer vision and pattern recognition, 2016;2818–2826.
36. Wang SH, and Zhang YD. DenseNet-201-based deep neural network with composite learning factor and precomputation for multiple sclerosis classification. ACM Transactions on Multimedia Computing, Communications, and Applications (TOMM), 2020;16(2).
37. Turkoglu M, Hanbay D, and Sengur A. Multi-model LSTM-based convolutional neural networks for detection of apple diseases and pests. J Ambient Intell Humaniz Comput, 2022;13(7):3335–3345.
38. Sandler M, Howard A, Zhu M, Zhmoginov A, and Chen LC. MobileNetV2: Inverted residuals and linear bottlenecks, 2018;4510–4520.
39. Pearson K. LIII. On lines and planes of closest fit to systems of points in space. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 1901; 2(11):559–572.
40. İmak A, Doğan G, Şengür A, and Ergen B. Asma Yaprağı türünün sınıflandırılması için doğal ve sentetik verilerden derin öznitelikler çıkarma, birleştirme ve seçmeye dayalı yeni bir yöntem. International Journal of Pure and Applied Sciences, 2023;9(1):46–55.
41. Cortes C, and Vapnik V. Support-vector networks. Mach Learn, 1995;20(3):273–297.
42. Serin J, Vidhya KT, Deepa IMI, Ebenezer V, and Jenefa A. Gender classification from fingerprint using hybrid CNN-SVM. Journal of Artificial Intelligence and Technology, (2024);4(1):82-87.
43. Demir F. DeepBreastNet: A novel and robust approach for automated breast cancer detection from histopathological images. Biocybernetics and biomedical engineering, (2021);41(3):1123-1139.
44. Polat H, Türkoğlu M, Polat O, and Şengür A. A novel approach for accurate detection of the DDoS attacks in SDN-based SCADA systems based on deep recurrent neural networks. Expert Systems with Applications, (2022);197:116748.
45. Doğan G, and Ergen B. A new mobile convolutional neural network-based approach for pixel-wise road surface crack detection. Measurement, 2022;195:111119.
46. Doğan G, and Ergen B. Karayollarındaki asfalt çatlaklarının tespiti için yeni bir konvolüsyonel sinir ağı tabanlı yöntem. Fırat University Journal of Engineering Science, 2022;34(2):485–494.
47. Imak A, Çelebi A, Polat O, Türkoğlu M, and Şengür A. ResMIBCU-Net: an encoder–decoder network with residual blocks, modified inverted residual block, and bi-directional ConvLSTM for impacted tooth segmentation in panoramic X-ray images. Oral Radiol., 2023;1:1–15.