Forecasting National Electricity Consumption Using Machine Learning Methods: An Application with UK Data

Abstract

Forecasting electricity demand is vital for operating energy systems in a secure, economical, and environmentally sustainable manner. Recently, studies on electricity demand forecasting methods have focused on improving classical methods and integrating new technologies. This study aimed to forecast electricity demand using the electricity demand dataset from the United Kingdom. For this purpose, we utilized machine learning methods such as Long Short-Term Memory (LSTM), eXtreme Gradient Boosting (XGBoost) and Boosted Linear Regression using the Python programming language in the Google Colab environment. We compared the method performances using accuracy measurement criteria such as Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). The results demonstrate the superiority of the LSTM method over the other methods.

Keywords

• Electricity Demand Forecasting
• LSTM
• XGBoost
• Boosted Linear Regression

ULUSAL ELEKTRİK TÜKETİMİNİN MAKİNE ÖĞRENMESİ YÖNTEMLERİYLE TAHMİN EDİLMESİ: BİRLEŞİK KRALLIK VERİLERİYLE BİR UYGULAMA

Abstract

Elektrik enerjisi talebinin tahmin edilmesi, enerji sistemlerinin güvenli, ekonomik ve çevresel olarak sürdürülebilir bir şekilde işletilmesi için hayati öneme sahiptir. Son zamanlarda, elektrik talep tahmin yöntemleri üzerine yapılan çalışmalar hem klasik yöntemlerin iyileştirilmesi hem de yeni teknolojilerin entegrasyonu üzerine yoğunlaşmıştır. Bu çalışmada, Birleşik Krallık'taki elektrik talebi veri seti kullanılarak elektrik talebinin tahmin edilmesi amaçlanmıştır. Bu amaç doğrultusunda, Uzun Kısa Süreli Bellek (LSTM), Aşırı Gradyan Artırma Algoritması (XGBoost) ve Lineer Artırma Regresyonu gibi çeşitli tahmin modelleri Google Colab ortamında Python yazılım dili kullanılarak uygulanmıştır. Modellerin performansları, Kök Ortalama Kare Hata (RMSE) ve Ortalama Mutlak Yüzde Hata (MAPE) doğruluk ölçüm yöntemleri kullanılarak karşılaştırılmıştır. Çalışmanın sonuçları LSTM yönteminin diğer yöntemlere üstünlüğünü göstermektedir.

Keywords

• Elektrik Talebi Tahmini
• LSTM
• XGBoost
• Artırımlı Lineer Regresyon

References

1. Aldarraji, M., Vega-Márquez, B., Pontes, B., Mahmood, B. ve Riquelme, J.C. (2024) Addressing energy challenges in Iraq: Forecasting power supply and demand using artificial intelligence models, Heliyon, 10(4), doi:10.1016/j.heliyon.2024.e25821
2. Arıkök, Y.A. (2024) A Comparative Study of ARIMA and LSTM for Two-Days Ahead Forecasting of Electricity Demand in İzmir-Manisa Region in Turkey, [Yayınlanmamış Yüksek Lisans Tezi], Yaşar University
3. Bashir, T., Haoyong, C., Tahir, M.F. ve Liqiang, Z. (2022) Short term electricity load forecasting using hybrid prophet-LSTM model optimized by BPNN, Energy Reports, 8: 1678-1686, doi:10.1016/j.egyr.2021.12.067
4. Chai, T. and Draxler, R.R. (2014) Root Mean Square Error (RMSE) or Mean Absolute Error (MAE)? - Arguments against Avoiding RMSE in the Literature. Geoscientific Model Development, 7, 1247-1250, doi:10.5194/gmd-7-1247-2014
5. de Myttenaere, A., Golden, B., Le Grand, B. ve Rossi, F. (2016). Mean Absolute Percentage Error for regression models. Neurocomputing, 192, 38-48. doi:10.1016/j.neucom.2015.12.114
6. Ensafi, Y., Amin, S.H., Zhang, G. ve Shah, B. (2022) Time-series forecasting of seasonal items sales using machine learning - A comparative analysis, International Journal of Information Management Data Insights, 2(1), doi:10.1016/j.jjimei.2022.100058
7. Farsi, B., Amayreh, A., Mazila, K. ve Chuah, J. H. (2021). Genetic algorithm and sparrow search algorithm based deep learning predictor for short-term PV power forecasting. IEEE Access, 9, 127259-127271. doi:10.1109/ACCESS.2021.311216
8. Grandón, G.T., Schwenzer, J., Steens, T. ve Breuing, J. (2024) Electricity demand forecasting with hybrid classical statistical and machine learning algorithms: Case study of Ukraine, Applied Energy, 355, doi:10.1016/j.apenergy.2023.122249

9. Gülay, E., Sen, M. ve Akgün, Ö.B. (2024) Forecasting electricity production from various energy sources in Türkiye: A predictive analysis of time series, deep learning, and hybrid models, Energy, 286, doi:10.1016/j.energy.2023.129566
10. Güven, D. ve Kayalıca, M.Ö. (2023) Analysing the determinants of the Turkish household electricity consumption using gradient boosting regression tree, Energy for Sustainable Development, 77, doi:10.1016/j.esd.2023.101312
11. Grzeszczyk, T.A. ve Grzeszczyk, M.K. (2023). Justifying Short-Term Load Forecasts Obtained with the Use of Neural Models. Energies, 15(5), 1852. doi:10.3390/en15051852
12. Ilić, I., Görgülü, B., Çevik, M. ve Baydoğan, M.G. (2021) Explainable boosted linear regression for time series forecasting, Pattern Recognition, 120, doi:10.1016/j.patcog.2021.108144
13. Khan, Z.A., Ullah, A., Haq, I.U., Hamdy, M., Mauro, G.M., Khan, M., Hijji, M. ve Baik, S.W. (2022) Efficient Short-Term Electricity Load Forecasting for Effective Energy Management, Sustainable Energy Technologies and Assessments, 53, Part A, doi:10.1016/j.seta.2022.102337
14. Lewis, C. D. (1982). Industrial and Business Forecasting Methods: A Practical Guide to Exponential Smoothing and Curve Fitting. Butterworth Scientific.
15. Li, C., Zhang, M., Förderer, K., Matthes, J. ve Hagenmeyer, V. (2024). A comparative analysis of machine learning algorithms for aggregated electric chargepoint load forecasting. E3S Web of Conferences, 545, 01004. doi:10.1051/e3sconf/202454501004
16. Lim, B., Arık, S. Ö., Loeff, N. ve Pfister, T. (2021). Temporal Fusion Transformers for interpretable multi-horizon time series forecasting. International Journal of Forecasting, 37(4), 1748-1764. doi:10.1016/j.ijforecast.2021.03.012
17. https://pypi.org/project/linear-tree, Erişim Tarihi: 06.12.2024, Konu: Linear Tree Kütüphanesi
18. Machlev, R., Heistrene, L., Perl, M., Levy, K., Belikov, J., Mannor, S. ve Levron, Y. (2022). Explainable Artificial Intelligence (XAI) techniques for energy and power systems: Review, challenges and opportunities. Energy and AI, 9, 100169. doi:10.1016/j.egyai.2022.100169
19. Massaoudi, M., Abu-Rub, H., Refaat, S. S., Chihi, I. ve Oueslati, F. S. (2021). Deep learning in smart grid technology: A review of recent advancements and future prospects. IEEE Access, 9, 54558-54584. doi:10.1109/ACCESS.2021.3071269
20. Nofal, S. (2023) Forecasting next-hour electricity demand in small-scale territories: Evidence from Jordan, Heliyon, 9(9), doi:10.1016/j.heliyon.2023.e19790
21. Oreshkin, B. N., Carpov, D., Chapados, N. ve Bengio, Y. (2020). N-BEATS: Neural basis expansion analysis for interpretable time series forecasting. International Conference on Learning Representations (ICLR), https://arxiv.org/abs/1905.10437
22. Rick, R. ve Berton, L. (2022) Energy forecasting model based on CNN-LSTM-AE for many time series with unequal lengths, Engineering Applications of Artificial Intelligence, 113, doi:10.1016/j.engappai.2022.104998
23. Sarwar, S., Aziz, G. ve Tiwari, A.K. (2023) Implication of machine learning techniques to forecast the electricity price and carbon emission: Evidence from a hot region, Geoscience Frontiers, 15(3), doi:10.1016/j.gsf.2023.101647
24. Sebastian, N. E. (2024) A Comprehensive Analysis of UK Electricity Consumption Using the OSI Model for Data Communication and Predictive Analytics, M.Sc. Research Project, National College of Ireland, 1-21.
25. Sha’aban, Y. A. (2024) Predictive models for short-term load forecasting in the UK’s electrical grid, PLOS ONE, 19(4), doi:10.1371/journal.pone.0297267
26. Su, X., Yan, X. ve Tsai, C-L. (2012) Linear regression, WIREs Computational Statistics, 4: 275-294. doi:10.1002/wics.1198
27. Waheed, W. ve Xu, Q., (2024) The COVID-19 pandemic's impact on user consumption: A shift towards very efficient load forecasting, Sustainable Energy, Grids and Networks, 39, doi:10.1016/j.segan.2024.101426
28. Wang, Y., Sun, S., Chen, X., Zeng, X., Kong, Y., Chen, J., Guo, Y. ve Wang, T. (2021) Short-term load forecasting of industrial customers based on SVMD and XGBoost. International Journal of Electrical Power & Energy Systems, 129, doi:10.1016/j.ijepes.2021.106830
29. Zhang, B., Li, N., Shi, F. ve Law, R. (2020) A deep learning approach for daily tourist flow forecasting with consumer search data, Asia Pacific Journal of Tourism Research, 25(3): 323-339, doi:10.1080/10941665.2019.1709876