Research on Predicting Wind Turbine Power Generation based on Hybrid LSTM-GNN Model

Junjie Liu; Juntao Liu

doi:10.6919/ICJE.202412_10(12).0008

Authors

Junjie Liu
Juntao Liu

DOI:

https://doi.org/10.6919/ICJE.202412_10(12).0008

Keywords:

Wind Power; LSTM; GNN; Spatiotemporal Data; Forecasting Model.

Abstract

Wind power, as a significant clean energy source, faces challenges related to intermittency and uncertainty in its forecasting. This paper presents a hybrid LSTM-GNN wind power forecasting method based on spatiotemporal data. The proposed method first utilizes LSTM to process time series data and extract temporal features. It then employs GNN to handle the graph-structured data of wind farms to extract spatial features. Finally, these two types of features are combined for wind power prediction. By integrating both temporal and spatial information, this approach offers more accurate wind power forecasts, which contributes to grid stability and energy management. Experimental results demonstrate that the LSTM-GNN hybrid model performs exceptionally well in wind turbine forecasting, with RMSE and MAE values aligning with expectations, indicating that the combined model effectively enhances prediction accuracy and learning efficiency.

Downloads

Download data is not yet available.

References

[1] Hochreiter, S. and J. Schmidhuber, Long Short-Term Memory. Neural Computation, 1997. 9(8): p. 1735-1780.

[2] Dolatabadi, A., H. Abdeltawab, and Y.A.R.I. Mohamed, Hybrid Deep Learning-Based Model for Wind Speed Forecasting Based on DWPT and Bidirectional LSTM Network. IEEE Access, 2020. 8: p. 229219-229232.

[3] Zhen, H., et al. A Hybrid Deep Learning Model and Comparison for Wind Power Forecasting Considering Temporal-Spatial Feature Extraction. Sustainability, 2020. 12, DOI: 10.3390/su12229490.

[4] Fan, H., et al. M2GSNet: Multi-Modal Multi-Task Graph Spatiotemporal Network for Ultra-Short-Term Wind Farm Cluster Power Prediction. Applied Sciences, 2020. 10, DOI: 10.3390/app10217915.

[5] Li, H., L. Liu, and Q. He, A Spatiotemporal Coupling Calculation-Based Short-Term Wind Farm Cluster Power Prediction Method. IEEE Access, 2023. 11: p. 131418-131434.

[6] Gilbert, C., J. Browell, and D. McMillan, Leveraging Turbine-Level Data for Improved Probabilistic Wind Power Forecasting. IEEE Transactions on Sustainable Energy, 2020. 11(3): p. 1152-1160.

[7] Madsen, H., et al., Standardizing the Performance Evaluation of Short-Term Wind Power Prediction Models. Wind Engineering, 2005. 29(6): p. 475-489.

[8] Al-Masri, H.M.K., A.A. Almehizia, and M. Ehsani, Accurate Wind Turbine Annual Energy Computation by Advanced Modeling. IEEE Transactions on Industry Applications, 2017. 53(3): p. 1761-1768.

[9] Peric, B., A. Simonovic, and M. Vorkapić, Comparative analysis of numerical computational techniques for determination of the wind turbine aerodynamic performances. Thermal Science, 2020. 25: p. 175-175.

[10] Grasso, F., Usage of Numerical Optimization in Wind Turbine Airfoil Design. Journal of Aircraft, 2011. 48: p. 248-255.

[11] Djavareshkian, m.H., J.A. Bidarouni, and M.R. Saber, New approach to high-fidelity aerodynamic design optimization of a wind turbine blade. International Journal of Renewable Energy Research, 2013. 3: p. 725-734.

[12] Scarselli, F., et al., The Graph Neural Network Model. IEEE Transactions on Neural Networks, 2009. 20(1): p. 61-80.

[13] Dou, J.Z.a.X.L.a.Y.X.a.J.S.a.J.L.a.Y.M.a.D., SDWPF: A Dataset for Spatial Dynamic Wind Power Forecasting Challenge at KDD Cup 2022. 2022.

[14] Miele, E.S., N. Ludwig, and A. Corsini Multi-Horizon Wind Power Forecasting Using Multi-Modal Spatio-Temporal Neural Networks. Energies, 2023. 16, DOI: 10.3390/en16083522.