Comparative Analysis of LSTM Architectures for Wind Speed Forecasting: A Case Study in Muş, Turkey

Abstract

This study assesses the effectiveness of five distinct Long Short-Term Memory (LSTM) architectures for forecasting wind speed in Muş, Turkey. The models include Vanilla LSTM, Stacked LSTM, Bidirectional LSTM, Attention LSTM, and Residual LSTM. The data, obtained from the Muş Meteorological Office, underwent preprocessing to handle missing values by averaging the same day and month values between 1969 and 2023. The dataset, containing 20,088 daily wind speed measurements, was split into training and test sets, with 80% allocated for training and 20% for testing. Each model was trained over 100 epochs with a batch size of 32, and performance was assessed using Mean Squared Error (MSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). The Vanilla LSTM model showed the lowest MSE and MAE values, indicating superior overall performance, while the Attention LSTM model achieved the lowest MAPE, demonstrating better percentage accuracy. These findings indicate that the Vanilla and Attention LSTM models are the most effective for wind speed forecasting, with the choice between them depending on the prioritization of total error versus percentage error.

Keywords

• LSTM
• Time series forecasting
• Energy
• Wind Speed
• Muş

References

1. Hochreiter S, Schmidhuber J. Long Short-Term Memory. Neural Comput 1997; 9: 1735–1780.
2. Governorship of Muş, www.mus.gov.tr.
3. Shang Z, He Z, Chen Y, et al. Short-term wind speed forecasting system based on multivariate time series and multi-objective optimization. Energy 2022; 238: 122024.
4. Wu C, Wang J, Chen X, et al. A novel hybrid system based on multi-objective optimization for wind speed forecasting. Renew Energy 2020; 146: 149–165.
5. Aly HHH. An intelligent hybrid model of neuro Wavelet, time series and Recurrent Kalman Filter for wind speed forecasting. Sustain Energy Technol Assessments 2020; 41: 100802.
6. Liu X, Lin Z, Feng Z. Short-term offshore wind speed forecast by seasonal ARIMA - A comparison against GRU and LSTM. Energy 2021; 227: 120492.
7. He J, Xu J. Ultra-short-term wind speed forecasting based on support vector machine with combined kernel function and similar data. EURASIP J Wirel Commun Netw 2019; 2019: 248.
8. Xu W, Ning L, Luo Y. Wind Speed Forecast Based on Post-Processing of Numerical Weather Predictions Using a Gradient Boosting Decision Tree Algorithm. Atmosphere (Basel) 2020; 11: 738.

9. Wang Y, Zou R, Liu F, et al. A review of wind speed and wind power forecasting with deep neural networks. Appl Energy 2021; 304: 117766.
10. Joseph LP, Deo RC, Prasad R, et al. Near real-time wind speed forecast model with bidirectional LSTM networks. Renew Energy 2023; 204: 39–58.
11. Köse B, Güneser MT. Assessment of Wind Characteristics and Wind Energy Potential in West Black Sea Region Of Turkey. Eskişehir Tech Univ J Sci Technol A - Appl Sci Eng 2019; 20: 227–237.
12. Wadi M, Kekezoglu B, Baysal M, et al. Feasibility Study of Wind Energy Potential in Turkey: Case Study of Catalca District in Istanbul. In: 2019 2nd International Conference on Smart Grid and Renewable Energy (SGRE). IEEE, pp. 1–6.
13. Onat N, Ersoz S. Analysis of wind climate and wind energy potential of regions in Turkey. Energy 2011; 36: 148–156.
14. Arslan H, Baltaci H, Akkoyunlu BO, et al. Wind speed variability and wind power potential over Turkey: Case studies for Çanakkale and İstanbul. Renew Energy 2020; 145: 1020–1032.
15. Sirdaş S. Daily wind speed harmonic analysis for Marmara region in Turkey. Energy Convers Manag 2005; 46: 1267–1277.
16. Shao B, Song D, Bian G, et al. Wind Speed Forecast Based on the LSTM Neural Network Optimized by the Firework Algorithm. Adv Mater Sci Eng 2021; 2021: 1–13.
17. Pradhan PP, Subudhi B. Wind speed forecasting based on wavelet transformation and recurrent neural network. Int J Numer Model Electron Networks, Devices Fields; 33. Epub ahead of print 12 January 2020. DOI: 10.1002/jnm.2670.
18. Lu P, Ye L, Zhao Y, et al. Review of meta-heuristic algorithms for wind power prediction: Methodologies, applications and challenges. Appl Energy 2021; 301: 117446.
19. Hu Y-L, Chen L. A nonlinear hybrid wind speed forecasting model using LSTM network, hysteretic ELM and Differential Evolution algorithm. Energy Convers Manag 2018; 173: 123–142.
20. Chen M-R, Zeng G-Q, Lu K-D, et al. A Two-Layer Nonlinear Combination Method for Short-Term Wind Speed Prediction Based on ELM, ENN, and LSTM. IEEE Internet Things J 2019; 6: 6997–7010.
21. Alhussan AA, M. El-Kenawy E-S, Abdelhamid AA, et al. Wind speed forecasting using optimized bidirectional LSTM based on dipper throated and genetic optimization algorithms. Front Energy Res; 11. Epub ahead of print 1 June 2023. DOI: 10.3389/fenrg.2023.1172176.
22. Chandra DR, Kumari MS, Sydulu M. A detailed literature review on wind forecasting. In: 2013 International Conference on Power, Energy and Control (ICPEC). 2013, pp. 630–634.
23. Keren B., Sabitha K. Probabilistic forecasting of wind power generation using extreme learning machineinvolving bootstrap method. ITECH 2016; 04: 729–736.
24. Azad HB, Mekhilef S, Ganapathy VG. Long-Term Wind Speed Forecasting and General Pattern Recognition Using Neural Networks. IEEE Trans Sustain Energy 2014; 5: 546–553.
25. Wang X, Guo P, Huang X. A Review of Wind Power Forecasting Models. Energy Procedia 2011; 12: 770–778.
26. Sharma N, Deo R. Wind speed forecasting in Nepal using self-organizing map-based online sequential extreme learning machine. In: Predictive Modelling for Energy Management and Power Systems Engineering. Elsevier, pp. 437–484.
27. Subramani K, J SS, Habelalmateen MI, et al. Predicting Wind Energy: Machine Learning from Daily Wind Data. E3S Web Conf 2024; 540: 03009.
28. Yang X, Delworth TL, Jia L, et al. Skillful seasonal prediction of wind energy resources in the contiguous United States. Commun Earth Environ 2024; 5: 313.
29. Tugal I, Sevgin F. Analysis and forecasting of temperature using time series forecasting methods a case study of Mus. Therm Sci 2023; 27: 3081–3088.
30. Lu J, Wang Y, Zhu Y, et al. DACLnet: A Dual-Attention-Mechanism CNN-LSTM Network for the Accurate Prediction of Nonlinear InSAR Deformation. Remote Sens 2024; 16: 2474.
31. Lattari F, Rucci A, Matteucci M. A Deep Learning Approach for Change Points Detection in InSAR Time Series. IEEE Trans Geosci Remote Sens 2022; 60: 1–16.
32. Wu Y, Yuan M, Dong S, et al. Remaining useful life estimation of engineered systems using vanilla LSTM neural networks. Neurocomputing 2018; 275: 167–179.
33. Ma M, Liu C, Wei R, et al. Predicting machine’s performance record using the stacked long short‐term memory (LSTM) neural networks. J Appl Clin Med Phys; 23. Epub ahead of print 16 March 2022. DOI: 10.1002/acm2.13558.
34. da Silva DG, Meneses AA de M. Comparing Long Short-Term Memory (LSTM) and bidirectional LSTM deep neural networks for power consumption prediction. Energy Reports 2023; 10: 3315–3334.
35. Ghosh S, Ekbal A, Bhattacharyya P. Natural language processing and sentiment analysis: perspectives from computational intelligence. In: Computational Intelligence Applications for Text and Sentiment Data Analysis. Elsevier, pp. 17–47.
36. Muhammad K, Mustaqeem, Ullah A, et al. Human action recognition using attention based LSTM network with dilated CNN features. Futur Gener Comput Syst 2021; 125: 820–830.
37. Kang Q, Chen EJ, Li Z-C, et al. Attention-based LSTM predictive model for the attitude and position of shield machine in tunneling. Undergr Sp 2023; 13: 335–350.
38. Wen X, Li W. Time Series Prediction Based on LSTM-Attention-LSTM Model. IEEE Access 2023; 11: 48322–48331.
39. Yue B, Fu J, Liang J. Residual Recurrent Neural Networks for Learning Sequential Representations. Information 2018; 9: 56.
40. Vatsa A, Hati AS, Kumar P, et al. Residual LSTM-based short duration forecasting of polarization current for effective assessment of transformers insulation. Sci Rep 2024; 14: 1369.
41. Fu S, Zhang Y, Lin L, et al. Deep residual LSTM with domain-invariance for remaining useful life prediction across domains. Reliab Eng Syst Saf 2021; 216: 108012.