Assessment of Electricity Generation Using Deep Learning on Solar Power Plants

Year 2024, , 289 - 311, 31.08.2024

Yunus Emre Kıymaz , Hidayet Oguz

https://doi.org/10.47112/neufmbd.2024.49

Abstract

Solar panel technology is expensive to install such systems, which have a lifespan of about 25 years on average. It is often important to make production estimates for the future to make optimal use of these systems. This study assesses three two-year daily frequency data sets and a one-year hourly frequency data set from the solar power plants (univariate time series) based in Konya, which have a 1MW capacity per annum. Electricity production analysis is conducted based on the data from the solar power plants using deep learning. The preferred method is determined to be Long Short-Term Memory (LSTM), and it has been compared with another statistical method used in time series analysis, Seasonal Autoregressive Integrated Moving Average (SARIMA). The results obtained with each dataset have been subjected to five different performance measurement mechanisms (MSE, RMSE, NMSE, MAE, MAPE and R2). It has been observed that the LSTM model generally provides results closer to real data compared to the SARIMA model. According to the RMSE score, the average value of four power plants is 973 in LSTM and 1361 in SARIMA, in this case LSTM gave a successful result compared to SARIMA. Before establishing a solar power plant, carrying out a feasibility study has a profitability-enhancing role.

Keywords

Deep Learning, Long Short-Term Memory, Seasonal Autoregressive Integrated Moving Average, Solar Power Plant, Univariate Time Series

Güneş Enerjisi Santrallerinde Derin Öğrenme Kullanılarak Elektrik Üretiminin Değerlendirilmesi

Abstract

Güneş paneli teknolojisi ortalama 25 yıl ömrü olan bu tür sistemlerin kurulumu pahalıdır. Bu sistemlerden en iyi şekilde yararlanmak için geleceğe yönelik üretim tahminleri yapmak çoğu zaman önemlidir. Bu çalışmada, Konya merkezli yıllık 1MW kapasiteye sahip güneş enerjisi santrallerine (tek değişkenli zaman serisi) ait iki yıllık üç günlük frekans veri seti ve bir yıllık saatlik frekans veri seti değerlendirilmektedir. Elektrik üretim analizi, derin öğrenme kullanılarak güneş enerjisi santrallerinden elde edilen verilere dayanılarak yapılmaktadır. Tercih edilen yöntem uzun kısa süreli hafıza (LSTM) olup, zaman serisi analizinde kullanılan diğer bir istatistiksel yöntem olan mevsimsel otoregresif bütünleşik hareketli ortalama (SARIMA) ile kıyaslanmıştır. Her bir veri seti ile elde edilmiş sonuçlar beş farklı performans ölçüm mekanizmasına (MSE, RMSE, NMSE, MAE, MAPE ve R2) tabi tutulmuş ve LSTM modelinin genellikle SARIMA modeline göre daha gerçek verilere yakın sonuçlar verdiği tespit edilmiştir. RMSE skoruna göre dört santralin ortalama değeri LSTM'de 973, SARIMA'da 1361 olup, bu durumda LSTM, SARIMA'ya göre başarılı bir sonuç vermiştir. Güneş enerjisi santrali kurmadan önce fizibilite çalışmasının yapılması karlılığı artırıcı bir role sahiptir.

Keywords

Derin Öğrenme, Güneş Enerjisi Santrali, Mevsimsel Otoregresif Bütünleşik Hareketli Ortalama, Uzun Kısa Süreli Hafıza, Tek Değişkenli Zaman Serileri

References

• A.L. Samuel, Some studies in machine learning using the game of checkers, IBM Journal of Research and Development. 3 (1959), 210-229. doi:10.1147/rd.33.0210.
• Y. LeCun, Y. Bengio, G. Hinton, Deep learning, Nature. 521 (2015), 436-444. doi:10.1038/nature14539.
• R. Roy, AI, ML, and DL: How not to get them mixed!, (2019). https://towardsdatascience.com/understanding-the-difference-between-ai-ml-and-dl-cceb63252a6c (access date 07 January 2021).
• H. Lütkepohl, M. Krätzig, P.C.B. Phillips, Applied time series econometrics, Cambridge University Press, 2004.
• D. Cano, J.M. Monget, M. Albuisson, H. Guillard, N. Regas, L. Wald, A method for the determination of the global solar radiation from meteorological satellite data, Solar Energy. 37 (1986), 31-39. doi:10.1016/0038-092X(86)90104-0.
• S.E. Rusen, Modeling and analysis of global and diffuse solar irradiation components using the satellite estimation method of HELIOSAT, Computer Modeling in Engineering & Sciences. 115 (2018), 327-343.
• S. Ener Rusen, A. Konuralp, Quality control of diffuse solar radiation component with satellite-based estimation methods, Renewable Energy. 145 (2020), 1772-1779. doi:10.1016/j.renene.2019.07.085.
• M. Abdel-Nasser, K. Mahmoud, Accurate photovoltaic power forecasting models using deep LSTM-RNN, Neural Computing and Applications. 31 (2019), 2727-2740. doi:10.1007/s00521-017-3225-z.
• R.K. Agrawal, F. Muchahary, M.M. Tripathi, Long term load forecasting with hourly predictions based on long-short-term-memory networks. In 2018 IEEE Texas Power and Energy Conference (TPEC), IEEE, 2018: ss. 1-6. doi:10.1109/TPEC.2018.8312088.
• S. Balluff, J. Bendfeld, S. Krauter, Short term wind and energy prediction for offshore wind farms using neural networks. In 2015 International Conference on Renewable Energy Research and Applications (ICRERA), IEEE, 2015: ss. 379-382. doi:10.1109/ICRERA.2015.7418440.
• A. Gensler, J. Henze, B. Sick, N. Raabe, Deep Learning for solar power forecasting — An approach using AutoEncoder and LSTM Neural Networks. In 2016 IEEE International Conference on Systems, Man, and Cybernetics (SMC), IEEE, 2016: ss. 002858-002865. doi:10.1109/SMC.2016.7844673.
• H. Sharadga, S. Hajimirza, R.S. Balog, Time series forecasting of solar power generation for large-scale photovoltaic plants, Renewable Energy. 150 (2020), 797-807. doi:10.1016/j.renene.2019.12.131.
• U. Şencan, Short term electricity price forecasting using Long Short-Term Memory, Thesis, Bahçeşehir University, 2018.
• F. Özen, R. Ortaç Kabaoğlu, T.V. Mumcu, Deep learning based temperature and humidity prediction, Necmettin Erbakan University Journal of Science and Engineering. (2023). doi:10.47112/neufmbd.2023.20.
• M. Hacibeyoglu, M. Çelik, Ö. Erdaş Çiçek, Energy efficiency estimation in buildings with K nearest neighbor algorithm, Necmettin Erbakan University Journal of Science and Engineering. 5 (2) (2023), 65-74. doi:10.47112/neufmbd.2023.10.
• N.C. Alparslan, A. Kayabasi, S.E. Rusen, Estimation of global solar radiation by using ANN and ANFIS. In 2019 Innovations in Intelligent Systems and Applications Conference (ASYU), IEEE, 2019: ss. 1-6. doi:10.1109/ASYU48272.2019.8946448.
• W. Donat, What is Python: An Intro to a Cross-Platform Programming Language, (2015). https://www.atlantic.net/vps-hosting/what-is-python-intro-cross-platform-programming-language/ (access date 01 June 2021).
• Anaconda Software Distribution, Anaconda Documentation. (2020). https://docs.anaconda.com/ (access date 01 October 2021).
• Anaconda Navigator, (2020). https://docs.anaconda.com/anaconda/navigator/ (access date 07 January 2021).
• C.R. Harris, K.J. Millman, S.J. van der Walt, R. Gommers, P. Virtanen, D. Cournapeau, E. Wieser, J. Taylor, S. Berg, N.J. Smith, R. Kern, M. Picus, S. Hoyer, M.H. van Kerkwijk, M. Brett, A. Haldane, J.F. del Río, M. Wiebe, P. Peterson, P. Gérard-Marchant, K. Sheppard, T. Reddy, W. Weckesser, H. Abbasi, C. Gohlke, T.E. Oliphant, Array programming with NumPy, Nature. 585 (2020), 357-362. doi:10.1038/s41586-020-2649-2.
• J.D. Hunter, Matplotlib: A 2D Graphics Environment, Computing in Science & Engineering. 9 (2007), 90-95. doi:10.1109/MCSE.2007.55.
• M. Abadi, A. Agarwal, P. Barham, E. Brevdo, Z. Chen, C. Citro, G.S. Corrado, A. Davis, J. Dean, M. Devin, S. Ghemawat, I. Goodfellow, A. Harp, G. Irving, M. Isard, Y. Jia, R. Jozefowicz, L. Kaiser, M. Kudlur, J. Levenberg, D. Mane, R. Monga, S. Moore, D. Murray, C. Olah, M. Schuster, J. Shlens, B. Steiner, I. Sutskever, K. Talwar, P. Tucker, V. Vanhoucke, V. Vasudevan, F. Viegas, O. Vinyals, P. Warden, M. Wattenberg, M. Wicke, Y. Yu, X. Zheng, TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems, (2016). https://arxiv.org/abs/1603.04467.
• M. Najibi, G. Lai, A. Kundu, Z. Lu, V. Rathod, T. Funkhouser, C. Pantofaru, D. Ross, L.S. Davis, A. Fathi, DOPS: Learning to Detect 3D Objects and Predict their 3D Shapes, Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. (2020), 11913-11922. http://arxiv.org/abs/2004.01170.
• W. McKinney, Data Structures for Statistical Computing in Python. In 2010: ss. 56-61. doi:10.25080/Majora-92bf1922-00a.
• F. Chollet, Keras: Deep learning for humans, (2015).
• S. Seabold, J. Perktold, Statsmodels: Econometric and Statistical Modeling with Python. In 2010: ss. 92-96. doi:10.25080/Majora-92bf1922-011.
• Turkish State Meteorological Service, Türkiye Global Güneş Radyasyonu Uzun Yıllar Ortalaması (2004-2018), (2018). https://www.mgm.gov.tr/kurumici/radyasyon_iller.aspx (access date 07 January 2021).
• S. Hochreiter, J. Schmidhuber, Long Short-Term Memory, Neural Computation. 9 (1997), 1735-1780. doi:10.1162/neco.1997.9.8.1735.
• C. Olah, Understanding LSTM Networks, (2015). http://colah.github.io/posts/2015-08-Understanding-LSTMs/ (access date 07 January 2021).
• G.E.P. Box, G.M. Jenkins, Time series analysis: forecasting and control, Holden-Day, 1970. https://books.google.com.tr/books?id=5BVfnXaq03oC.
• M. Ghofrani, M. Alolayan, Time Series and Renewable Energy Forecasting. In Time Series Analysis and Applications, InTech, 2018: ss. 77-92. doi:10.5772/intechopen.70845.
• Ö. Zeydan, Zonguldak bölgesi pm10 konsantrasyonu dağılımının modellenmesi, Thesis, Kocaeli Üniversitesi, 2014. http://dspace.kocaeli.edu.tr:8080/xmlui/handle/11493/856.
• Anonymous, MSE, RMSE, MAE, MAPE ve Diğer Metrikler, (2017). https://veribilimcisi.com/2017/07/14/mse-rmse-mae-mape-metrikleri-nedir/ (access date 07 January 2021).