드론 촬영 이미지와 딥러닝 기법을 이용한 태양광 패널의 먼지오염 탐지
Detection of Photovoltaic Panel Dust Using Drone Shooting Imagery and Deep Learning Techniques
0
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
태양광 모듈 표면의 먼지오염은 수광장애 요인으로 발전량 저하를 유발하기 때문에 이를 주기적으로 탐지하는 것은 유지관리 측면에서 매우 중요하다. 따라서 본 연구에서는 드론으로 촬영한 이미지와 딥러닝 기법을 이용하여 태양광 패널 규모에서의 먼지오염 여부에 대한 분류 예측 모델의 성능을 평가하였다. 강원특별자치도 삼척시에 위치한 육상 및 옥상 태양광 패널에 대한 드론 이미지를 취득 및 전처리 후 5종의 딥러닝 기법을 적용하여 패널위의 먼지오염 여부를 분류하였으며 혼동행렬을 이용하여 예측결과의 분류 성능을 평가하였다. 그 결과 convolutional neural network 기법이 가장 높은 예측력을 보였으며 Area under the curve는 80.6%, 정확도는 66.0%, 재현율은 94.0%로계산되었다. 실제로 패널위에 먼지오염이 있는 경우에 높은 예측력을 보인 반면 먼지오염이 없는경우를 오분류한 비율도 높았기에 이에 대한 개선이 필요할 것으로 판단된다. 본 연구에서 제안한분석모델은 대규모 태양광 발전소에서 패널위의 먼지오염을 신속하게 평가하고 스크리닝하는 방법론의 하나로 활용될 수 있을 것으로 기대한다.
Abstract
Regular detection of dust on photovoltaic (PV) modules is crucial for their maintenance because dust can hinder light harvesting and decrease power generation. The current study utilized drone images of panels and deep learning techniques to detect dust and evaluate the performance of a prediction model. After acquiring and preprocessing the drone images of the panels in Samcheok city, Gangwon state, we applied five deep learning techniques to identify and categorize the panels as dusty or clean. We evaluated the performance of these techniques using a confusion matrix. The convolutional neural network technique demonstrated the highest predictive performance with an area under the curve of 80.6%, classification accuracy of 66.0%, and recall of 94.0%. However, the model's high predictive performance for dusty panel cases was accompanied by a high rate of misclassification for the clean panel cases, indicating the need for improvement. The proposed analytical model can be used to quickly screen and evaluate the dusty panels in large-scale power plants.
참고문헌 (27)
- Almagro Armenteros, J. J., "DeepLoc: prediction of protein subcellular localization using deep learning", Bioinformatics, 33(21), pp.3387~ (2017)
- Chen, J., "Study on impacts of dust accumulation and rainfall on PV power reduction in East China", Energy, 194, pp.116915~ (2020)
- Cipriani, G., "Convolutional Neural Network for Dust and Hotspot Classification in PV Modules", Energies, 13(23), pp.6357~ (2020)
- Conceição, R., "Saharan dust transport to Europe and its impact on photovoltaic performance:A case study of soiling in Portugal", Solar Energy, 160, pp.94~102 (2018)
- Cruz-Rojas, T., "A novel comparison of image semantic segmentation techniques for detecting dust in photovoltaic panels using machine learning and deep learning", Renewable Energy, 217, pp.119126~ (2023)
- Gao, Y., "A deep learning-based method detects dust from solar PV panels through Unmaned Aerial Vehicles", Journal of Physics: Conference Series, 2584, pp.012019~ (2023)
- Godec, P., "Democratized image analytics by visual programming through integration of deep models and small-scale machine learning", Nature communications, 10(1), pp.4551~ (2019)
- Han, J., "A Study on the Contamination of Photovoltaic Cells by Fine Dust in the Air", Transactions of Korean Hydrogen and New Energy Society, 29(3), pp.292~298 (2018)
- Jung, D., "Development of Machine Learning Models for Predicting Air Overpressure in an Open-pit Mine", Journal of The Korean Society of Mineral and Energy Resources Engineers, 59(1), pp.59~68 (2022)
- Kaldellis, J. K., "Systematic experimental study of the pollution deposition impact on the energy yield of photovoltaic installations", Renewable energy, 36(10), pp.2717~2724 (2011)
- Kim, S. K., "Detection of Surface Contamination of Photovoltaic Module Using Image Sensor", Proceeding of 46th the Korean Institute of Electrical Engineers, pp.1086~1087 (2015)
- Kim, D., "Machine Learning-Based Solar Power Efficiency Prediction Scheme Considering PM10 Fine-Dust Data", Proceedings of 23rd the Korean Institute of Information Scientists and Engineers, pp.272~274 (2021)
- Krizhevsky, A., "Imagenet classification with deep convolutional neural networks", Advances in neural information processing systems, pp.25~ (2012)
- Lee, G., "A Case Study for Analyzing the Optimal Location for A Solar Power Plant via AHP Analysis with Fine Dust and Weather", Journal of the Korea Safety Management & Science, 19(4), pp.157~167 (2017)
- Lobo, J. M., "AUC:a misleading measure of the performance of predictive distribution models", Global ecology and Biogeography, 17(2), pp.145~151 (2008)
- Menoufi, K., "Dust accumulation on photovoltaic panels: a case study at the East Bank of the Nile (Beni-Suef, Egypt)", Energy Procedia, 128, pp.24~31 (2017)
- Park, S., "The Prediction of Photovoltaic Power Using Regression Models Based on Weather Big-data and Sensing Data", The Transactions of the Korean Institute of Electrical Engineers, 68(12), pp.1662~1668 (2019)
- Rao, A., "Influence of Dust Deposition on Photovoltaic Panel Performance", Energy Procedia, 54, pp.690~700 (2014)
- "Renewable Energy Cloud Platform" (2023)[URL]
- Saini, R. K., "Effects of dust on the performance of solar panels – a review update from 2015-2020", Energy & Environment, 34(6), pp.2110~2162 (2023)
- IEA-PVPS task 13: Performance, operation and reliability of photovoltaic systems–Soiling Losses – Impact on the Performance of Photovoltaic Power Plants, "Schill, C.", International Energy Agency, ~ (2022)
- Son, J., "A Research of the Effects of Particulate Matter on Solar Photovoltaic", Proceedings of Korea Environmental Policy, pp.52~54 (2019)
- Sung, S., "Prediction of Photovoltaic Power Generation Based on Machine Learning Considering the Influence of Particulate Matter", Environmental and Resource Economics Review, 28(4), pp.467~495 (2019)
- Szegedy, C., "Rethinking the inception architecture for computer vision", IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.2818~2826 (2016)
- You, H., "Deep neural network photovoltaic power generation forecasting with fine dust data", Proceeding of 51st the Korean Institute of Electrical Engineers, pp.547~548 (2020)