臺灣博碩士論文加值系統

現今太陽能光電廠的建置容量越來越大。在太陽光電廠的尋檢過程中，通常會使用空拍機來拍攝整座太陽光電廠。然而，在進行拍攝之前，需要進行繁雜的前置作業，且即使拍攝完成後，仍需要人工檢查太陽能板是否正常運作。因此，本論文提出了利用手持式紅外線相機結合視覺辨識的方法，協助案場尋檢人員更快速地確認太陽本次研究基於物件偵測模型之深度學習的太陽能板熱斑檢測系統，將熱顯像照比對出溫度異常處，如有異常把可見光照利用模型圈選出太陽能板位置並將太陽能板編碼。之後將熱顯像照跟可見光照比對溫度異常處是否為太陽能板，最後生成試算表，顯示出檢驗成果。

The installed capacity of solar power plants is increasing nowadays, however, the maintaince of the Solar panel becomes important. Drones are usually used to capture images of the entire facility inorder o maintain the quality of the installed capacity through the inspection process of solar power plants. However, shooting schedule of the preparatory work are complex, manual visual inspection is needed to ensure the normal power output of solar panels after the shooting. In this study, we proposed a method with using handheld infrared camera and visual recognition to assist inspection personnel in quickly confirming the condition of solar panels. This research is based on a deep learning solar panel hotspot detection system using the YOLOv5 object detection model. By comparing thermal imaging photos to detect temperature abnormalities, the system can circle the spots of the solar panel onto the photograph and encode them. The following procedure of the system compares the thermal imaging photos and visible light photos to makesure whether the temperature abnormalities are caused by the solar panel itself.

摘要......i
Abstract......ii
誌謝......iii
目錄......iv
表目錄......vi
圖目錄......vii
第一章緒論......1
1.1前言......1
1.2研究動機與目的......2
第二章 文獻回顧......3
2.1太陽光電發展......3
2.1.1矽晶太陽能板......3
2.1.2太陽光電發電廠現況......4
2.1.3太陽光電維運現況......5
2.1.4太陽光電板熱斑形成原因......6
2.1.5 減少熱班產生......7
2.2 熱斑形成熱顯像圖像分析......8
2.3 熱像儀......12
2.4影像辨識......14
第三章 實驗方法......27
3.1研究流程圖......27
3.2資料收集......28
3.3 熱斑模擬......30
3.4圖片標記......31
3.4.1YOLO資料標記格式......32
3.5熱顯像照灰階數值提取......33
3.6 YOLO v5 模型訓練......35
3.7評估方法......36
3.7.1 mAP......36
3.7.2 IoU......37
3.8 照片座標資料......38
第四章 結果與討論......39
4.1 實驗設備及處理平台......39
4.2 樣本數與相關實驗資料準備......41
4.3 訓練結果......41
4.4 物件辨識結果......42
4.5 論文實驗結果......43
第五章 結論......45
參考文獻......46
Abstract......49
Experimental......50
Results and discussion......50
Conclusions......52

[1]Shockley, W., The theory of p-n junctions in semiconductors and p-n junction transistors. The Bell System Technical Journal, 1949. 28(3): p. 435-489.
[2]Wolf, S.D., et al., High-efficiency Silicon Heterojunction Solar Cells: A Review. 2012. 2(1): p. 7-24.
[3]Binetti, S., et al., Key success factors and future perspective of silicon-based solar cells. 2013. 2013.
[4]Chaturvedi, P., B. Hoex, and T.M. Walsh, Broken metal fingers in silicon wafer solar cells and PV modules. Solar Energy Materials and Solar Cells, 2013. 108: p. 78-81.
[5]Dhimish, M., et al., The impact of cracks on photovoltaic power performance. Journal of Science: Advanced Materials and Devices, 2017. 2(2): p. 199-209.
[6]Sulaiman, S.A., et al., Influence of Dirt Accumulation on Performance of PV Panels. Energy Procedia, 2014. 50: p. 50-56.
[7]Zaihidee, F.M., et al., Dust as an unalterable deteriorative factor affecting PV panel's efficiency: Why and how. Renewable and Sustainable Energy Reviews, 2016. 65: p. 1267-1278.
[8]Gade, R. and T.B. Moeslund, Thermal cameras and applications: a survey. Machine Vision and Applications, 2014. 25(1): p. 245-262.
[9]Chevrette, P.J.A.o.P.T., An Overview of Thermal Imagers and Their Basic Components. 1995: p. 223-228.
[10]Venkataraman, B., B.J.I.-N.-D.T. Raj, and C. Monitoring, Performance parameters for thermal imaging systems. 2003. 45(8): p. 531-535.
[11]Szafron, C.J.W.U.o.T.P., Application of thermal imaging in electrical equipment examination. 2014.
[12]Jian, M., et al., Saliency detection using quaternionic distance based weber local descriptor and level priors. Multimedia Tools and Applications, 2018. 77(11): p. 14343-14360.
[13]Uijlings, J.R.R., et al., Selective Search for Object Recognition. International Journal of Computer Vision, 2013. 104(2): p. 154-171.
[14]Zitnick, C.L. and P. Dollár. Edge Boxes: Locating Object Proposals from Edges. in Computer Vision – ECCV 2014. 2014. Cham: Springer International Publishing.
[15]Ren, S., et al., Faster r-cnn: Towards real-time object detection with region proposal networks. 2015. 28.
[16]Redmon, J., et al. You Only Look Once: Unified, Real-Time Object Detection. in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2016.
[17]Liu, W., et al. Ssd: Single shot multibox detector. in Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14. 2016. Springer.
[18]Liu, W., et al. SSD: Single Shot MultiBox Detector. in Computer Vision – ECCV 2016. 2016. Cham: Springer International Publishing.
[19]Lin, J., et al. An Asynchronous Convolution Process Engine forVGG-16 Neural Network. in 2020 IEEE International Conference on Integrated Circuits, Technologies and Applications (ICTA). 2020.
[20]Zhang, H., X.J.M.T. Hong, and Applications, Recent progresses on object detection: a brief review. 2019. 78: p. 27809-27847.
[21]Vincent, L. Morphological grayscale reconstruction: definition, efficient algorithm and applications in image analysis. in Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. 1992.
[22]Studio encoding parameters of digital television for standard 4:3 and wide-screen 16:9 aspect ratios. 2007.
[23]Dasiopoulou, S., et al., Knowledge-assisted semantic video object detection. IEEE Transactions on Circuits and Systems for Video Technology, 2005. 15(10): p. 1210-1224.
[24]Redmon, J., et al., You Only Look Once: Unified, Real-Time Object Detection. 2015: p. 779-788.
[25]Redmon, J., et al. You only look once: Unified, real-time object detection. in Proceedings of the IEEE conference on computer vision and pattern recognition. 2016.
[26]Everingham, M., et al., The pascal visual object classes challenge: A retrospective. 2015. 111: p. 98-136.
[27]Wu, W., et al., Application of local fully Convolutional Neural Network combined with YOLO v5 algorithm in small target detection of remote sensing image. 2021. 16(10): p. e0259283.
[28]Wan, C., et al., Overview of YOLO Object Detection Algorithm. 2022.
[29]Čuljak, I., et al., A brief introduction to OpenCV. 2012: p. 1725-1730.
[30]Dubois, P.F. Introduction to Python. 1996.
[31]Samimi, H.J.J.o.C.S.i.C., Introduction to the Python programming language. 2013. 29: p. 8-9.
[32]Abadi, M., et al., TensorFlow: A system for large-scale machine learning. 2016. abs/1605.08695.
[33]Albawi, S., et al., Understanding of a convolutional neural network. 2017: p. 1-6.
[34]O’Shea, K. and R.J.A. Nash, An Introduction to Convolutional Neural Networks. 2015. abs/1511.08458.
[35]Sakib, S., et al. An Overview of Convolutional Neural Network: Its Architecture and Applications. 2018.
[36]Hinton, G.E., et al., Improving neural networks by preventing co-adaptation of feature detectors. 2012.
[37]LeCun, Y., et al., Backpropagation applied to handwritten zip code recognition. 1989. 1(4): p. 541-551.
[38]Albawi, S., T.A. Mohammed, and S. Al-Zawi. Understanding of a convolutional neural network. in 2017 International Conference on Engineering and Technology (ICET). 2017.
[39]LeCun, Y., et al., Handwritten digit recognition with a back-propagation network. 1989. 2.
[40]Nair, V. and G.E. Hinton. Rectified linear units improve restricted boltzmann machines. in Proceedings of the 27th international conference on machine learning (ICML-10). 2010.
[41]Arif, R.B., et al. Study and observation of the variations of accuracies for handwritten digits recognition with various hidden layers and epochs using convolutional neural network. in 2018 4th International Conference on Electrical Engineering and Information & Communication Technology (iCEEiCT). 2018. IEEE.
[42]Linardatos, P., V. Papastefanopoulos, and S.J.E. Kotsiantis, Explainable ai: A review of machine learning interpretability methods. 2020. 23(1): p. 18.
[43]Chowdhery, A., et al., Palm: Scaling language modeling with pathways. 2022.
[44]Bray, T., The javascript object notation (json) data interchange format. 2014.
[45]Gan, K., et al., Artificial intelligence detection of distal radius fractures: a comparison between the convolutional neural network and professional assessments. 2019. 90(4): p. 394-400.