LSTM ve GRU Ağları Kullanılarak Türkiye’nin Elektrik Enerjisi Tüketiminin Tahmin Edilmesi

Öz

Enerji talep yönetimi, gelişmekte olan ve yükselen ekonomiler için özellikle önemlidir. Büyüyen ekonomilerine bağlı olarak enerji tüketimleri önemli ölçüde artmaktadır. Türkiye’nin hızlı ekonomik ve nüfus artışının bir sonucu olarak elektrik tüketimi artmaktadır. Elektrik tüketimi tahmini enerji tedarikçileri, tüketiciler ve politika yapıcılar için önemli bir rol oynar. Bu nedenle, gelecekteki elektrik tüketim eğilimlerini doğru ve güvenilir bir şekilde tahmin etmek için modellerin kullanılması, elektrik güç sistemlerinin planlanması ve işletilmesi için kilit bir konudur. Bu makale, zaman serisi verileri için Uzun Kısa-Süreli Bellek (Long Short-Term Memory-LSTM) ve Kapılı Yinelemeli Birim (Gated Recurrent Unit-GRU) modelleri gibi derin öğrenme yöntemlerini kullanarak elektrik enerjisi tüketimini tahmin etmeye odaklanmıştır. Türkiye’de elektrik enerjisi tüketiminin geçmişe dönük veri seti kullanılarak bir saatlik ve üç saatlik ileriye yönelik tahminler gerçekleştirilmiştir. Karşılaştırma sonuçları, GRU modelinin LSTM modelinden biraz daha iyi olduğunu göstermektedir. Çalışmamız ayrıca, bir saat ileri tahminlerin üç saat ileri tahminlerden daha doğru olduğunu ortaya koymaktadır.

Anahtar Kelimeler

• Derin Öğrenme
• Elektrik Enerji Tüketimi
• Tahmin
• Kapılı Yinelemeli Birim
• Uzun Kısa-Süreli Bellek

Forecasting of Turkey’s Electrical Energy Consumption using LSTM and GRU Networks

Öz

Energy demand management is particularly important for developing and emerging economies. Their energy consumptions increase significantly, depending on their growing economies. As a result of Turkey’s rapid economic and population growth, electricity consumption is increasing. Electricity consumption forecasting plays an essential role for energy suppliers, consumers, and policy makers. Therefore, using models to accurately and reliably forecast future electricity consumption trends is a key issue for the planning and operation of electric power systems. This paper focused on forecasting electrical energy consumption by utilizing deep learning methods, i.e., Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models, for time series data. One-hour and three-hour ahead forecasting are accomplished by using a historical dataset of electrical energy consumption in Turkey. The comparison results show that the GRU model is slightly better than that of the LSTM. Our study also reveals that one-hour ahead predictions are more accurate than three-hour ahead predictions.

Anahtar Kelimeler

• Deep Learning
• Electrical Energy Consumption
• Forecasting
• Gated Recurrent Unit
• Long Short-Term Memory

Kaynakça

1. Koç, E., & Şenel, M. C. (2013). Dünyada ve Türkiye’de enerji durumu–genel değerlendirme. Mühendis ve Makina Dergisi, 54(639), 32-44.
2. Bilgili, M. (2010). Present status and future projections of electrical energy in Turkey.Gazi University Journal of Science, 23(2), 237-248.
3. Fackrell, B. (2013). Turkey and regional energy Security on the road to 2023. Turkish Policy Quarterly, 12(2), 83-89.
4. TUIK, Turkish Statistics Institute. (2021). Statistics, http://www.tuik.gov.tr.
5. Turkey's Lessons for Emerging Economies - Caixin Global. http://www.caixinglobal.com,(20.01.2021).
6. International Monetary Fund (2021). World Economic Outlook Database, October 2020. https://www.imf.org/en/home, (20.01.2021).
7. World Bank (2021), International Comparison Program database: GDP, PPP (current international $). https://data.worldbank.org/indicator/NY.GDP.MKTP.PP.CD, (20.01.2021).
8. The World Factbook (2021), Real GDP (purchasing power parity). https://www.cia.gov/the-world-factbook/field/real-gdp-purchasing-power-parity/, (23.01.2021).

9. World Data (2021). Turkey Energy Consumption. https://www.worlddata.info/asia/turkey/energy-consumption.php, (23.01.2021).
10. De Felice, M., Alessandri, A., & Ruti, P. M. (2013). Electricity demand forecasting over Italy: Potential benefits using numerical weather prediction models. Electric Power Systems Research, 104, 71-79.
11. Kavaklioglu, K., Ceylan, H., Ozturk, H. K., & Canyurt, O.E. (2009). Modeling and prediction of Turkey’s electricity consumption using artificial neural networks. Energy Conversion and Management, 50(11), 2719-2727.
12. Hamzaçebi, C., & Kutay, F. (2004).Yapay sinir ağları ile Türkiye elektrik enerjisi tüketiminin 2010 yılına kadar tahmini. Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi, 19(3), 227-233.
13. Dilaver, Z., & Hunt, L. C. (2011). Turkish aggregate electricity demand: An outlook to 2020. Energy, 36(11), 6686-6696.
14. Bolturk, E., Oztaysi, B., & Sari, I. U. (2012). Electricity Consumption Forecasting Using Fuzzy Time Series. IEEE Symposium on Computational Intelligence and Informatics. 20-22 November, Budapest, Hungary, 245-249.
15. Tokgöz, A., & Ünal, G. (2018). A RNN Based Time Series Approach for Forecasting Turkish Electricity Load.26th Signal Processing and Communications Applications Conference (SIU).2-5 May, Izmir, Turkey, 1-4.
16. Akay, D., & Atak, M. (2007). Grey prediction with rolling mechanism for electricity demand forecasting of Turkey. Energy, 32(9), 1670-1675.
17. Hamzacebi, C., & Es, H. A. (2014). Forecasting the annual electricity consumption of Turkey using an optimized grey model. Energy, 70,165-171.
18. Hu, Y. C. (2017). Electricity consumption prediction using a neural-network-based grey forecasting approach.Journal of the Operational Research Society, 68(10), 1259-1264.
19. Yumurtacı, Z., & Asmaz, E. (2004). Electric energy demand of Turkey for the year 2050. Energy Sour, 26(12), 1157-1164.
20. Kavaklioglu, K. (2011). Modeling and prediction of Turkey’s electricity consumption using support vector regression. Applied Energy, 88(1), 368-375.
21. Oğcu, G., Demirel, O.F., & Zaim, S. (2012). Forecasting electricity consumption with neural networks and support vector regression. Social and Behavioral Sciences, 58, 1576-1585.
22. Kavaklioglu, K. (2014). Robust electricity consumption modeling of Turkey using singular value decomposition. International Journal of Electrical Power & Energy Systems, 54, 268-276.
23. Karaca, C., & Karacan, H. (2016). Çoklu regresyon metoduyla elektrik tüketim talebini etkileyen faktörlerin incelenmesi. Selçuk Üniversitesi Mühendislik, Bilim ve Teknoloji Dergisi, 4(3), 182-195.
24. Haliloğlu, E. Y., & Tutu, B. E. (2018). Türkiye için kısa vadeli elektrik enerjisi talep tahmini. Journal of Yaşar University, 13(51), 243-255.
25. Topalli, A. K., Erkmen, I., & Topalli, I. (2006). Intelligent short-term load forecasting in Turkey. International Journal of Electrical Power & Energy Systems, 28(7), 437-447.
26. Erdogdu, E. (2007). Electricity demand analysis using cointegration and ARIMA modelling: A case study of Turkey. Energy Policy, 35(2), 1129-1146.
27. Demirel, Ö., Kakilli, A., & Tektaş, M. (2010). ANFIS ve ARMA modelleri ile elektrik enerjisi yük tahmini. Gazi Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi, 25(3), 601-610.
28. Boran, K. (2014). The Box Jenkins approach to forecast net electricity consumption in Turkey. Energy Sour A, 36(5), 515-524.
29. Çevik, H. H., & Çunkaş, M. (2015) Short-term load forecasting using fuzzy logic and ANFIS. Neural Computing and Applications, 26(6), 1355-1367.
30. Ozturk, H. K., Ceylan, H., Canyurt, O. E., & Hepbasli, A. (2005). Electricity estimation using genetic algorithm approach: a case study of Turkey. Energy, 30(7), 1003-1012.
31. Yiğit, V. (2011). Genetik algoritma ile Türkiye net elektrik enerjisi tüketiminin 2020 yılına kadar tahmini. International Journal of Engineering Research and Development, 3(2), 37-41.
32. Kucukali, S., & Baris, K. (2010). Turkey’s short-term gross annual electricity demand forecast by fuzzy logic approach. Energy Policy, 38(5), 2438-2445.
33. Toksarı, M. D. (2009). Estimating the net electricity energy generation and demand using the ant colony optimization approach: case of Turkey. Energy Policy, 37, 1181-1187.
34. Kıran, M. S., Özceylan, E., Gündüz, M., & Paksoy, T. (2012). Swarm intelligence approaches to estimate electricity energy demand in Turkey. Knowledge-Based Systems, 36, 93-103.
35. Kavaklioglu, K., Ceylan, H., Ozturk, H. K., & Canyurt, O. E. (2009). Modeling and prediction of Turkey’s electricity consumption using artificial neural networks. Energy Conversion and Management, 50(11), 2719-2727.
36. Çunkaş, M., & Altun, A. A. (2010). Long term electricity demand forecasting in Turkey using artificial neural networks. Energy Sour B, 5(3), 279-289.
37. Sözen, A., Isikan, O., Menlik, T., & Arcaklioglu, E. (2011). The forecasting of net electricity consumption of the consumer groups in Turkey. Energy Sour B, 6, 20-46.
38. Yetis, Y., & Jamshidi, M. (2014). Forecasting of Turkey’s Electricity Consumption using Artificial Neural Network. World Automation Congress (WAC).3-7 August, Waikoloa, USA, 723-728.
39. Günay, M. E. (2016). Forecasting annual gross electricity demand by artificial neural networks using predicted values of socio-economic indicators and climatic conditions: Case of Turkey. Energy Policy, 90, 92-101.
40. Hamzaçebi, C., Es, H. A., & Çakmak, R. (2019). Forecasting of Turkey’s monthly electricity demand by seasonal artificial neural network. Neural Computing and Applications, 31(7), 2217-2231.
41. Özkurt, N., Öztura, H. Ş., & Güzeliş, C. (2020). 24-hour Electricity Consumption Forecasting for Day ahead Market with Long Short-Term Memory Deep Learning Model. 12th International Conference on Electrical and Electronics Engineering (ELECO). 26-28 November, Bursa, Turkey, 173-177.
42. Özbay, H., & Dalcali, A. (2021). Effects of COVID-19 on electric energy consumption in Turkey and ANN-based short-term forecasting. Turkish Journal of Electrical Engineering & Computer Sciences, 29(1), 78-97.
43. Schuster, M., & Paliwal, K. (1997). Bidirectional recurrent neural networks. IEEE Transactions on Signal Processing, 45(11), 2673-2681.
44. Koutnik, J., Greff, K., Gomez, F., & Schmidhuber, J. (2014). A Clockwork RNN. 31st International Conference on Machine Learning.21-26 June, Beijing, China, 1863-1871.
45. Cho, K., Van Merrienboer, B., Bahdanau, D., & Bengio, Y. (2014). On the Properties of Neural Machine Translation: Encoder–Decoder Approaches. 8th Workshop on Syntax, Semantics and Structure in Statistical Translation (SSST-8), 25 October, Doha, Qatar, 103-111.
46. Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735-1780.
47. Türkünoğlu, A. (2019). Short Term Electricity Consumption Forecasting using Long Short-Term Memory Cells. Yüksek Lisans Tezi, İstanbul Teknik Üniversitesi, Enerji Enstitüsü, İstanbul.
48. Shahid, F., Zameer, A. & Muneeb, M. (2020). Predictions for Covid-19 with deep learning models of LSTM, GRU and Bi-LSTM. Chaos, Solitons & Fractals, 140, 1-9.
49. Gers, F. A., Schmidhuber, J., & Cummins, F. (2000). Learning to forget: Continual prediction with LSTM.Neural Computation, 12(10), 2451-2471.
50. Zeroual, A., Harrou, F., Dairi, A. & Sun, Y. (2020). Deep learning methods forforecasting Covid-19 time-series data: A comparative study. Chaos, Solitons & Fractals, 140, 1-12.
51. Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2014). Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling. NIPS 2014 Workshop on Deep Learning, 1-9.
52. EPİAŞ Şeffaflık Platformu. Türkiye Gerçek Zamanlı Elektrik Tüketim Verileri, https://seffaflik.epias.com.tr/transparency/tuketim/gerceklesen-tuketim/gercek-zamanli-tuketim.xhtml, (15.01.2021).