Spatio-temporal wind speed forecasting using graph networks and novel Transformer architectures

• Published in: Applied energy Vol. 333; p. 120565
• Main Authors: Bentsen, Lars Ødegaard, Warakagoda, Narada Dilp, Stenbro, Roy, Engelstad, Paal
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.03.2023
• Subjects: Graph neural networks; Multi-step; Spatio-temporal wind forecasting; Transformers; Transformers; Graph neural networks; Spatio-temporal wind forecasting; Multi-step
• ISSN: 0306-2619
• DOI: 10.1016/j.apenergy.2022.120565

To improve the security and reliability of wind energy production, short-term forecasting has become of utmost importance. This study focuses on multi-step spatio-temporal wind speed forecasting for the Norwegian continental shelf. In particular, the study considers 14 offshore measurement stations and aims to leverage spatial dependencies through the relative physical location of different stations to improve local wind forecasts and simultaneously output different forecasts for each of the 14 locations. Our multi-step forecasting models produce either 10-minute, 1- or 4-hour forecasts, with 10-minute resolution, meaning that the models produce more informative time series for predicted future trends. A graph neural network (GNN) architecture was used to extract spatial dependencies, with different update functions to learn temporal correlations. These update functions were implemented using different neural network architectures. One such architecture, the Transformer, has become increasingly popular for sequence modelling in recent years. Various alterations have been proposed to better facilitate time series forecasting, of which this study focused on the Informer, LogSparse Transformer and Autoformer. This is the first time the LogSparse Transformer and Autoformer have been applied to wind forecasting and the first time any of these or the Informer have been formulated in a spatio-temporal setting for wind forecasting. By comparing against spatio-temporal Long Short-Term Memory (LSTM) and Multi-Layer Perceptron (MLP) models, the study showed that the models using the altered Transformer architectures as update functions in GNNs were able to outperform these. Furthermore, we propose the Fast Fourier Transformer (FFTransformer), which is a novel Transformer architecture based on signal decomposition and consists of two separate streams that analyse the trend and periodic components separately. The FFTransformer and Autoformer were found to achieve superior results for the 10-minute and 1-hour ahead forecasts, with the FFTransformer significantly outperforming all other models for the 4-hour ahead forecasts. Our code to implement the different models are made publicly available at: https://github.com/LarsBentsen/FFTransformer. [Display omitted] •Generic spatio-temporal multi-step forecasting framework using Graph Networks.•Models are shown to improve wind forecasts by leveraging spatial information.•Show the effectiveness of various Transformer architectures for forecasting.•FFTransformer, an altered Transformer architecture using signal decomposition.•Show the competitive performance of our proposed FFTransformer.