Regional wind power forecasting based on CNN-BiLSTM-reliefF method
Regional wind power forecasting based on CNN-BiLSTM-reliefF method
The noisy and stochastic nature of wind power production introduces many challenges to the grid. Wind power forecasting methods can effectively predict these uncertainties and ensure the stable operation of power systems. Due to the superior performance of deep learning approaches in predicting nonlinear data, a hybrid method based on two-dimensional convolutional neural network (Conv2D), bidirectional long short-term memory (BiLSTM), and ReliefF algorithm is presented in this study. First, wind power data are preprocessed. Then, the processed data are entered into each Conv2D and BiLSTM network for temporal-spatial feature extraction. Finally, the features that are extracted from these two networks are combined by assigning weights to them through the ReliefF algorithm for final wind power prediction. The forecasting performance of the presented method is tested on a real aggregated dataset of Belgium for 15 minutes ahead regional wind power forecasting. According to the results, the proposed model has higher accuracy than long short-term memory, gated recurrent unit (GRU), BiLSTM, and Conv2D networks.
Bidirectional long short-term memory, Convolutional neural network, ReliefF, Wind power forecasting
Dehghan, Fariba
e0863ef3-0a6c-467f-87ed-824cbd16408c
Parsa Moghaddam, Mohsen
fade89c8-c47d-497a-b52f-8631de238f0d
Imani, Maryam
7175b856-1aa2-4b2d-9c30-74c7fd30ae07
30 July 2024
Dehghan, Fariba
e0863ef3-0a6c-467f-87ed-824cbd16408c
Parsa Moghaddam, Mohsen
fade89c8-c47d-497a-b52f-8631de238f0d
Imani, Maryam
7175b856-1aa2-4b2d-9c30-74c7fd30ae07
Dehghan, Fariba, Parsa Moghaddam, Mohsen and Imani, Maryam
(2024)
Regional wind power forecasting based on CNN-BiLSTM-reliefF method.
In 2024 11th Iranian Conference on Renewable Energy and Distribution Generation (ICREDG).
IEEE..
(doi:10.1109/ICREDG61679.2024.10607820).
Record type:
Conference or Workshop Item
(Paper)
Abstract
The noisy and stochastic nature of wind power production introduces many challenges to the grid. Wind power forecasting methods can effectively predict these uncertainties and ensure the stable operation of power systems. Due to the superior performance of deep learning approaches in predicting nonlinear data, a hybrid method based on two-dimensional convolutional neural network (Conv2D), bidirectional long short-term memory (BiLSTM), and ReliefF algorithm is presented in this study. First, wind power data are preprocessed. Then, the processed data are entered into each Conv2D and BiLSTM network for temporal-spatial feature extraction. Finally, the features that are extracted from these two networks are combined by assigning weights to them through the ReliefF algorithm for final wind power prediction. The forecasting performance of the presented method is tested on a real aggregated dataset of Belgium for 15 minutes ahead regional wind power forecasting. According to the results, the proposed model has higher accuracy than long short-term memory, gated recurrent unit (GRU), BiLSTM, and Conv2D networks.
This record has no associated files available for download.
More information
Published date: 30 July 2024
Keywords:
Bidirectional long short-term memory, Convolutional neural network, ReliefF, Wind power forecasting
Identifiers
Local EPrints ID: 505446
URI: http://eprints.soton.ac.uk/id/eprint/505446
PURE UUID: 2b08fc1d-ad19-4ef5-a42a-c08b2e2a92da
Catalogue record
Date deposited: 08 Oct 2025 16:55
Last modified: 09 Oct 2025 02:22
Export record
Altmetrics
Contributors
Author:
Fariba Dehghan
Author:
Mohsen Parsa Moghaddam
Author:
Maryam Imani
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics