研究生: |
黃彥植 Yen-Chih Huang |
---|---|
論文名稱: |
太陽能電廠故障之熱影像深度學習診斷及定位研究 Research on Thermography Deep Learning Diagnosis and Allocation for Solar Power Plant Faults |
指導教授: |
郭政謙
Cheng-Chien Kuo |
口試委員: |
黃維澤
Wei-Tzer Huang 張宏展 Hong-Chan Chang 吳瑞南 Ruay-Nan Wu 陳鴻誠 Hung-Cheng Chen 李俊耀 Chun-Yao Lee 張建國 Chien-Kuo Chang |
學位類別: |
博士 Doctor |
系所名稱: |
電資學院 - 電機工程系 Department of Electrical Engineering |
論文出版年: | 2023 |
畢業學年度: | 111 |
語文別: | 中文 |
論文頁數: | 118 |
中文關鍵詞: | 太陽能電廠 、熱斑故障 、紅外線熱影像 、深度學習 、無人機 |
外文關鍵詞: | Solar Power Plants, Hot Spot Failures, Thermal Imaging, Deep Learning, Unmanned Aerial Vehicle(UAV) |
相關次數: | 點閱:234 下載:0 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
再生能源需求上升,提高建置大型太陽能發電廠數量及面積,其管理與診斷技術就必須更加提升檢測速度與準確度。目前針對太陽能板常見的熱斑故障檢測方式為透過熱影像儀拍攝,由於太陽能電廠面積很大,不易透過人工方式檢測,因此採用具備熱影像相機的無人機進行拍攝,以監測太陽能板的健康狀態。
本研究提出一種方法由無人機收集影像資料,並利用深度學習方法進行熱斑故障診斷,且結合地理位置資訊系統提供故障定位服務。首先,透過無人機的可見光及熱影像相機採集整個太陽能電廠之影像資料,藉由正射影像技術及影像處理將太陽能面板從影像分割作為資料集。再來,利用提出之故障疊加法及生成對抗網路ProGAN來增強深度學習所需的訓練影像資料,改善訓練集資料不平衡問題。再者,透過實際案場資料作為測試集,以捲積神經網路Resnet-50進行熱斑故障之熱特徵訓練及測試分析,驗證本研究提出的方法能有效解決訓練數據不足及可實際應用之可行性。最後,透過圖資系統將故障面板標示在電廠可見光俯視圖之上。
本研究於實際太陽能電廠進行實測,達到由無人機採集影像資料後,透過系統平台上傳至異地伺服器資料庫,再經由伺服器進行影像處理及深度學習之辨識,並將結果顯示在太陽能電廠監控系統平台,以便尋找故障面板做更換。該方法可對於大型太陽能電廠在故障診斷與電廠效能評估上,提供一個具有價值的參考依據。
As the demand for renewable energy rises, the number and area of large-scale solar power plants will increase, and their management and diagnostic technologies must further improve detection speed and accuracy. Currently, the common method for hot spot fault detection of solar panels is to take pictures with a thermal imager. Due to the large area of the solar power station, manual inspection is not easy. Therefore, UAVs(unmanned aerial vehicles) equipped with thermal imaging cameras will be used to photograph the health of the solar panels.
This study proposes a method for diagnosing hotspot faults using image data collected by UAVs using deep learning methods and combined with geographic information system to provide fault location services. First, the image data of the entire solar power plant is collected through the visible light and thermal imager of the UAVs, and the solar panels are segmented from the image as a data set through orthophoto technology and image processing. In addition, the proposed Fault Overlay Method and the generated adversarial network ProGAN are used to enhance the training image data required for deep learning and solve the data imbalance problem in the training set. Furthermore, the actual case field data is used as the test set. The thermal feature training and test analysis of hot spot faults are carried out using the convolutional neural network Resnet-50. To verify the feasibility of the method proposed in this study to effectively solve the shortage of training data and be practically applicable. Finally, the fault panel is marked on the perspective view of the plant through the geographic information system.
In this study, field measurements were carried out in real solar power plants. After the UAV collects the data, it is uploaded to the remote server database through the system platform. The server performs image processing and deep learning recognition. Display the results in the solar plant monitoring system platform to find faulty panels for replacement. This research can provide a valuable reference for large-scale solar power plants in fault diagnosis and plant performance evaluation.
[1] IEC, "IEC 62446-2:2020, Photovoltaic (PV) systems - Requirements for testing, documentation and maintenance - Part 2: Grid connected systems - Maintenance of PV systems," 2020.
[2] H. A. Walker, Best Practices for Operation and Maintenance of Photovoltaic and Energy Storage Systems; 3rd Edition, United States, 2018.
[3] S. Europe, Operation & Maintenance Best Practices Guidelines (Version 5.0), December 2021.
[4] V. S. B. Kurukuru, A. Haque, and M. A. Khan, "Fault Classification for Photovoltaic modules using Thermography and Image Processing." pp. 1-6.
[5] S. An, M. Lee, S. Park, H. Yang, and J. So, “An ensemble of simple convolutional neural network models for MNIST digit recognition,” arXiv preprint arXiv:2008.10400, 2020.
[6] Y. Pan, X. Zhang, G. Cervone, and L. Yang, “Detection of Asphalt Pavement Potholes and Cracks Based on the Unmanned Aerial Vehicle Multispectral Imagery,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 11, no. 10, pp. 3701-3712, 2018.
[7] Y. Tao, M. Xu, Y. Zhong, and Y. Cheng, “GAN-Assisted Two-Stream Neural Network for High-Resolution Remote Sensing Image Classification,” Remote Sensing, vol. 9, no. 12, pp. 1328, 2017.
[8] D. Zhou, F. Tian, X. Tian, L. Sun, X. Huang, F. Zhao, N. Zhou, Z. Chen, Q. Zhang, M. Yang, Y. Yang, X. Guo, Z. Li, J. Liu, J. Wang, J. Wang, B. Wang, G. Zhang, B. Sun, W. Zhang, D. Kong, K. Chen, and X. Li, “Diagnostic evaluation of a deep learning model for optical diagnosis of colorectal cancer,” Nature Communications, vol. 11, no. 1, pp. 2961, 2020/06/11, 2020.
[9] M. Doniyorjon, R. Madinakhon, M. Shakhnoza, and Y.-I. Cho, “An Improved Method of Polyp Detection Using Custom YOLOv4-Tiny,” Applied Sciences, vol. 12, no. 21, pp. 10856, 2022.
[10] S.-D. Lu, M.-H. Wang, S.-E. Wei, H.-D. Liu, and C.-C. Wu, “Photovoltaic Module Fault Detection Based on a Convolutional Neural Network,” Processes, vol. 9, no. 9, pp. 1635, 2021.
[11] S. A. Zaki, H. Zhu, M. A. Fakih, A. R. Sayed, and J. Yao, “Deep-learning–based method for faults classification of PV system,” IET Renewable Power Generation, vol. 15, no. 1, pp. 193-205, 2021.
[12] A. Rico Espinosa, M. Bressan, and L. F. Giraldo, “Failure signature classification in solar photovoltaic plants using RGB images and convolutional neural networks,” Renewable Energy, vol. 162, pp. 249-256, 2020/12/01/, 2020.
[13] M. W. Akram, G. Li, Y. Jin, X. Chen, C. Zhu, X. Zhao, A. Khaliq, M. Faheem, and A. Ahmad, “CNN based automatic detection of photovoltaic cell defects in electroluminescence images,” Energy, vol. 189, pp. 116319, 2019/12/15/, 2019.
[14] P. D. Moskowitz, and V. M. Fthenakis, “Toxic materials released from photovoltaic modules during fires: health risks,” Solar cells, vol. 29, no. 1, pp. 63-71, 1990.
[15] L. Koester, S. Lindig, A. Louwen, A. Astigarraga, G. Manzolini, and D. Moser, “Review of photovoltaic module degradation, field inspection techniques and techno-economic assessment,” Renewable and Sustainable Energy Reviews, vol. 165, pp. 112616, 2022/09/01/, 2022.
[16] IEC, "IEC 61730-1:2016, Photovoltaic (PV) module safety qualification - Part 1: Requirements for construction," 2016.
[17] U. Standard, "Standard for Flat-Plate Photovoltaic Modules and Panels," 2002.
[18] S. N. A. M. Ghazali, M. Z. Sujod, and M. S. Jadin, "Forensic of Solar PV: A Review of Potential Faults and Maintenance Strategies." pp. 1-6.
[19] M. Davarifar, A. Rabhi, A. El-Hajjaji, and M. Dahmane, "Real-time model base fault diagnosis of PV panels using statistical signal processing." pp. 599-604.
[20] Y. Hu, J. Zhang, W. Cao, J. Wu, G. Y. Tian, S. J. Finney, and J. L. Kirtley, “Online Two-Section PV Array Fault Diagnosis With Optimized Voltage Sensor Locations,” IEEE Transactions on Industrial Electronics, vol. 62, no. 11, pp. 7237-7246, 2015.
[21] A. Livera, M. Theristis, G. Makrides, and G. E. Georghiou, “Recent advances in failure diagnosis techniques based on performance data analysis for grid-connected photovoltaic systems,” Renewable Energy, vol. 133, pp. 126-143, 2019/04/01/, 2019.
[22] M. Ma, H. Liu, Z. Zhang, P. Yun, and F. Liu, “Rapid diagnosis of hot spot failure of crystalline silicon PV module based on I-V curve,” Microelectronics Reliability, vol. 100-101, pp. 113402, 2019/09/01/, 2019.
[23] T. Trupke, B. Mitchell, J. W. Weber, W. McMillan, R. A. Bardos, and R. Kroeze, “Photoluminescence Imaging for Photovoltaic Applications,” Energy Procedia, vol. 15, pp. 135-146, 2012/01/01/, 2012.
[24] Z. Wu, Y. Hu, J. X. Wen, F. Zhou, and X. Ye, “A Review for Solar Panel Fire Accident Prevention in Large-Scale PV Applications,” IEEE Access, vol. 8, pp. 132466-132480, 2020.
[25] K. A. Kim, G. S. Seo, B. H. Cho, and P. T. Krein, “Photovoltaic Hot-Spot Detection for Solar Panel Substrings Using AC Parameter Characterization,” IEEE Transactions on Power Electronics, vol. 31, no. 2, pp. 1121-1130, 2016.
[26] K. A. Kim, and P. T. Krein, "Photovoltaic hot spot analysis for cells with various reverse-bias characteristics through electrical and thermal simulation." pp. 1-8.
[27] Y. Hu, W. Cao, J. Ma, S. J. Finney, and D. Li, “Identifying PV Module Mismatch Faults by a Thermography-Based Temperature Distribution Analysis,” IEEE Transactions on Device and Materials Reliability, vol. 14, no. 4, pp. 951-960, 2014.
[28] J. H. Wohlgemuth, and S. R. Kurtz, "How can we make PV modules safer?." pp. 003162-003165.
[29] D. Hausamann, W. Zirnig, G. Schreier, and P. Strobl, “Monitoring of gas pipelines – a civil UAV application,” Aircraft Engineering and Aerospace Technology, vol. 77, no. 5, pp. 352-360, 2005.
[30] A. Constantin, and R. N. Dinculescu, "UAV development and impact in the power system." pp. 1-5.
[31] F. Grimaccia, S. Leva, A. Dolara, and M. Aghaei, “Survey on PV Modules’ Common Faults After an O&M Flight Extensive Campaign Over Different Plants in Italy,” IEEE Journal of Photovoltaics, vol. 7, no. 3, pp. 810-816, 2017.
[32] F. P. García Márquez, and I. Segovia Ramírez, “Condition monitoring system for solar power plants with radiometric and thermographic sensors embedded in unmanned aerial vehicles,” Measurement, vol. 139, pp. 152-162, 2019/06/01/, 2019.
[33] F. Hong, J. Song, H. Meng, R. Wang, F. Fang, and G. Zhang, “A novel framework on intelligent detection for module defects of PV plant combining the visible and infrared images,” Solar Energy, vol. 236, pp. 406-416, 2022/04/01/, 2022.
[34] A. Greco, C. Pironti, A. Saggese, M. Vento, and V. Vigilante, "A deep learning based approach for detecting panels in photovoltaic plants." pp. 1-7.
[35] I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y. Bengio, “Generative adversarial networks,” Communications of the ACM, vol. 63, no. 11, pp. 139-144, 2020.
[36] A. Radford, L. Metz, and S. Chintala, “Unsupervised representation learning with deep convolutional generative adversarial networks,” arXiv preprint arXiv:1511.06434, 2015.
[37] M. Arjovsky, and L. Bottou, “Towards principled methods for training generative adversarial networks,” arXiv preprint arXiv:1701.04862, 2017.
[38] M. Arjovsky, S. Chintala, and L. Bottou, "Wasserstein generative adversarial networks." pp. 214-223.
[39] I. Gulrajani, F. Ahmed, M. Arjovsky, V. Dumoulin, and A. C. Courville, “Improved training of wasserstein gans,” Advances in neural information processing systems, vol. 30, 2017.
[40] Q. Yang, P. Yan, Y. Zhang, H. Yu, Y. Shi, X. Mou, M. K. Kalra, Y. Zhang, L. Sun, and G. Wang, “Low-Dose CT Image Denoising Using a Generative Adversarial Network With Wasserstein Distance and Perceptual Loss,” IEEE Transactions on Medical Imaging, vol. 37, no. 6, pp. 1348-1357, 2018.
[41] A. Osokin, A. Chessel, R. E. C. Salas, and F. Vaggi, "GANs for Biological Image Synthesis." pp. 2252-2261.
[42] B. Tang, Y. Tu, Z. Zhang, and Y. Lin, “Digital Signal Modulation Classification With Data Augmentation Using Generative Adversarial Nets in Cognitive Radio Networks,” IEEE Access, vol. 6, pp. 15713-15722, 2018.
[43] L. Chen, Q. Yang, and W. Yan, "Generative Adversarial Network based Data Augmentation for PV Module Defect Pattern Analysis." pp. 8422-8427.
[44] W. Zhang, Y. Luo, Y. Zhang, and D. Srinivasan, “SolarGAN: Multivariate Solar Data Imputation Using Generative Adversarial Network,” IEEE Transactions on Sustainable Energy, vol. 12, no. 1, pp. 743-746, 2021.
[45] W. Tang, Q. Yang, K. Xiong, and W. Yan, “Deep learning based automatic defect identification of photovoltaic module using electroluminescence images,” Solar Energy, vol. 201, pp. 453-460, 2020/05/01/, 2020.
[46] F. Lu, R. Niu, Z. Zhang, L. Guo, and J. Chen, “A generative adversarial network-based fault detection approach for photovoltaic panel,” Applied Sciences, vol. 12, no. 4, pp. 1789, 2022.
[47] T. Karras, T. Aila, S. Laine, and J. Lehtinen, “Progressive growing of gans for improved quality, stability, and variation,” arXiv preprint arXiv:1710.10196, 2017.
[48] Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-based learning applied to document recognition,” Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, 1998.
[49] G. Cipriani, A. D’Amico, S. Guarino, D. Manno, M. Traverso, and V. Di Dio, “Convolutional Neural Network for Dust and Hotspot Classification in PV Modules,” Energies, vol. 13, no. 23, pp. 6357, 2020.
[50] R. H. Fonseca Alves, G. A. d. Deus Júnior, E. G. Marra, and R. P. Lemos, “Automatic fault classification in photovoltaic modules using Convolutional Neural Networks,” Renewable Energy, vol. 179, pp. 502-516, 2021/12/01/, 2021.
[51] D. Korkmaz, and H. Acikgoz, “An efficient fault classification method in solar photovoltaic modules using transfer learning and multi-scale convolutional neural network,” Engineering Applications of Artificial Intelligence, vol. 113, pp. 104959, 2022/08/01/, 2022.
[52] K. He, X. Zhang, S. Ren, and J. Sun, "Deep Residual Learning for Image Recognition." pp. 770-778.