Ampirik Kip Ayrıştırma Yöntemi ile Elde Edilen İçsel Kip Fonksiyonlarının Derin Öğrenme Tabanlı Rüzgâr Hızı Tahmin Modeli Başarımına Etkisinin Belirlenmesi

Öz

Son yıllarda küresel anlamda etkisini yoğun şekilde gösteren iklim değişikliğinin temelinde fosil yakıt tüketimi kaynaklı sera etkisinin kuvvetlenmesi yer aldığı bilinmektedir. İklim değişikliği neticesinde canlılar için hayati önem taşıyan su kaynaklarının azalacağı, ekolojik dengenin bozularak çölleşme ve kuraklığın artacağı öngörülmektedir. Bu sorunla başa çıkılabilmesi için fosil yakıt tüketiminin azaltılması ve enerji ihtiyacının yenilenebilir enerji kaynakları ile karşılanması gerekmektedir. Bu nedenle, temiz ve yenilenebilir bir enerji türü olan rüzgâr enerjisine olan ilgi dünya çapında her geçen gün artmaktadır. Bununla birlikte, rüzgâr hızının güçlü rastgeleliği ve durağan olmaması rüzgâr gücünün elektrik şebekesine entegre edilmesini zorlaştırmaktadır. Bu zorlukların üstesinden gelebilmek için rüzgâr hızının güvenilir ve yüksek doğrulukla tahmin edilmesi kritik önem arz etmektedir. Bu çalışmada, doğrusal olmayan dinamiklere sahip rüzgâr hızının yüksek doğrulukla tahmin edilebilmesi için ampirik kip ayrıştırma ve derin öğrenme yöntemlerinden uzun-kısa süreli bellek tekniklerini içeren melez modeldeki içsel kip fonksiyonlarının rüzgâr hızı tahmin performansı üzerindeki etkileri incelenmektedir. Türkiye’nin en yüksek rüzgâr enerji potansiyeline sahip bölgeleri arasında yer alan Marmara bölgesindeki Bandırma meteoroloji istasyonundan toplanan rüzgâr hızı verileri ampirik kip ayrıştırma tekniği ile içsel kip fonksiyonlarına ayrıştırılmaktadır. Her bir içsel kip fonksiyonunun tahmin modeli üzerindeki başarımının belirlenebilmesi için sırasıyla her bir içsel kip fonksiyonu derin öğrenme modeline dahil edilmeden tahmin modellerinin performansları ölçülmektedir. Tahmin modellerinin başarımları istatistiksel performans metriklerine göre hesaplanmaktadır.

Anahtar Kelimeler

• Ampirik kip ayrıştırma
• İçsel kip fonksiyonu
• Rüzgâr hızı tahmini
• Derin öğrenme
• Uzun-kısa süreli bellek.

Determining the Effect of Intrinsic Mode Functions Obtained by the Empirical Mode Decomposition on the Performance of Deep Learning based Wind Speed Prediction Model

Abstract

In recent years, it is known that the strengthening of the greenhouse effect caused by fossil fuel consumption is at the root of climate change, which has had an intense impact on the global scale. As a result of climate change, it is predicted that water resources, which are vital for living things, will decrease, and the ecological balance will deteriorate and desertification and drought will increase. In order to cope with this problem, fossil fuel consumption should be reduced and energy needs should be met with renewable energy sources. Therefore, the interest in wind energy, which is a clean and renewable energy type, is increasing day by day around the world. However, the strong randomness and non-stationarity of the wind speed make it difficult to integrate wind power into the electricity grid. To overcome these challenges, reliable and highly accurate estimation of wind speed is critical. In this study, the effects of intrinsic mode functions in the hybrid model, which includes empirical mode decomposition and long-short-term memory (LSTM) techniques, on wind speed prediction performance are investigated for high accuracy prediction of wind speed, which has nonlinear dynamics. Wind speed data collected from Bandırma meteorology station in the Marmara region, which is among the regions with the highest wind energy potential in Turkey, are decomposed to intrinsic mode functions by empirical mode decomposition technique. To determine the performance of each intrinsic mode function on the wind speed prediction model, the performances of the prediction models are measured without including each intrinsic mode function in the deep learning model, respectively. The performance of prediction models is computed according to statistical performance metrics.

Keywords

• Empirical mode decomposition
• Intrinsic mode function
• Wind speed prediction
• Deep learning
• Long-short term memory (LSTM)

Kaynakça

1. Akçay, H. & Filik, T. (2017). Short-term wind speed forecasting by spectral analysis from long-term observations with missing values. Applied Energy, 191, 653-662.
2. Altan, A. & Karasu, S. (2021). Ayrıştırma yöntemlerinin derin öğrenme algoritması ile tanımlanan rüzgâr hızı tahmin modeli başarımına etkisinin incelenmesi. Avrupa Bilim ve Teknoloji Dergisi, 20, 844-853.
3. Altan, A., Karasu, S., & Zio, E. (2021). A new hybrid model for wind speed forecasting combining long short-term memory neural network, decomposition methods and grey wolf optimizer. Applied Soft Computing, 100, 106996.
4. Cadenas, E., Rivera, W., Campos-Amezcua, R., & Heard, C. (2016). Wind speed prediction using a univariate ARIMA model and a multivariate NARX model. Energies, 9(2), 109.
5. Chen, C. F., Lai, M. C., & Yeh, C. C. (2012). Forecasting tourism demand based on empirical mode decomposition and neural network. Knowledge-Based Systems, 26, 281-287.
6. Chen, Y., Dong, Z., Wang, Y., Su, J., Han, Z., Zhou, D., Zhang, K., Zhao, Y., & Bao, Y. (2021). Short-term wind speed predicting framework based on EEMD-GA-LSTM method under large scaled wind history. Energy Conversion and Management, 227, 113559.
7. Chen, Y., He, Z., Shang, Z., Li, C., Li, L., & Xu, M. (2019). A novel combined model based on echo state network form multi-step ahead wind speed forecasting: A case study of NREL. Energy Conversion and Management, 179, 13-29.
8. Gauterin, E., Kammerer, P., Kühn, M., & Schulte, H. (2016). Effective wind speed estimation: comparison between Kalman filter and Takagi–Sugeno observer techniques. ISA Transactions, 62, 60-72.

9. Hochreiter, S. & Schmidhuber, J. (1997). Long short-term memory, neural computation, 9(8), 1735-1780.
10. Hoolohan, V., Tomlin, A. S., & Cockerill, T. (2018). Improved near surface wind speed predictions using Gaussian process regression combined with numerical weather predictions and observed meteorological data. Renewable Energy, 126, 1043-1054.
11. Huang, N. E., Shen, Z., Long, S. R., Wu, M. C., Shih, H. H., Zheng, Q., Yen, N. C, Tung, C. C., & Liu, H. H. (1998). The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis. Proceedings of the Royal Society of London, Series A: Mathematical, Physical and Engineering Sciences, 454(1971), 903-995.
12. Huang, N. E., Wu, M. L. C., Long, S. R., Shen, S. S., Qu, W., Gloersen, P., & Fan, K. L. (2003). A confidence limit for the empirical mode decomposition and Hilbert spectral analysis. Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences, 459(2037), 2317-2345.
13. Karasu, S., Altan, A., Saraç, Z., & Hacıoğlu, R. (2017a). Estimation of fast varied wind speed based on NARX neural network by using curve fitting. International Journal of Energy Applications and Technologies, 4(3), 137-146.
14. Karasu, S., Altan, A., Saraç, Z., & Hacıoğlu, R. (2017b). Prediction of wind speed with non-linear autoregressive (NAR) neural networks. IEEE 25th Signal Processing and Communications Applications Conference, Antalya-Turkey.
15. Liu, H. & Chen, C. (2019). Data processing strategies in wind energy forecasting models and applications: a comprehensive review. Applied Energy, 249, 392-408.
16. Liu, H., Duan, Z., Wu, H., Li, Y., & Dong, S. (2019). Wind speed forecasting models based on data decomposition, feature selection and group method of data handling network. Measurement, 148, 106971.
17. Liu, H., Tian, H. Q., & Li, Y. F. (2012). Comparison of two new ARIMA-ANN and ARIMA-Kalman hybrid methods for wind speed prediction. Applied Energy, 98, 415-424.
18. Liu, X., Lin, Z., & Feng, Z. (2021). Short-term offshore wind speed forecast by seasonal ARIMA-A comparison against GRU and LSTM. Energy, 227, 120492.
19. Ma, X., Jin, Y., & Dong, Q. (2017). A generalized dynamic fuzzy neural network based on singular spectrum analysis optimized by brain storm optimization for short-term wind speed forecasting. Applied Soft Computing, 54, 296-312.
20. Ma, Z., Chen, H., Wang, J., Yang, X., Yan, R., Jia, J., & Xu, W. (2020). Application of hybrid model based on double decomposition, error correction and deep learning in short-term wind speed prediction. Energy Conversion and Management, 205, 112345.
21. Ruiz-Aguilar, J. J., Turias, I., González-Enrique, J., Urda, D., & Elizondo, D. (2021). A permutation entropy-based EMD–ANN forecasting ensemble approach for wind speed prediction. Neural Computing and Applications, 33(7), 2369-2391.
22. Sainath, T. N., Vinyals, O., Senior, A., and Sak, H. (2015). Convolutional, long short-term memory, fully connected deep neural networks. IEEE International Conference on Acoustics, Speech and Signal Processing, Brisbane-Australia, 4580-4584.
23. Statistical Review of World Energy 2021. Available from: https://www.bp.com/content/dam/bp/business-sites /en/global/corporate/pdfs/energy-economics/statistical-re view/bp-stats-review-2021-renewable-energy.pdf
24. U.S. Energy Information Administration. Available from: https://www.eia.gov/renewable/data.php#wind.
25. Yan, X., Liu, Y., Xu, Y., & Jia, M. (2020). Multistep forecasting for diurnal wind speed based on hybrid deep learning model with improved singular spectrum decomposition. Energy Conversion and Management, 225, 113456.
26. Yu, C., Li, Y., Bao, Y., Tang, H., & Zhai, G. (2018). A novel framework for wind speed prediction based on recurrent neural networks and support vector machine. Energy Conversion and Management, 178, 137-145.
27. Zhang, D., Xu, Z., Li, C., Yang, R., Shahidehpour, M., Wu, Q., & Yan, M. (2019). Economic and sustainability promises of wind energy considering the impacts of climate change and vulnerabilities to extreme conditions. The Electricity Journal, 32(6), 7-12.
28. Zhao, Y., Ye, L., Pinson, P., Tang, Y., & Lu, P. (2018). Correlation-constrained and sparsity-controlled vector autoregressive model for spatio-temporal wind power forecasting. IEEE Transactions on Power Systems, 33(5), 5029-5040.