Issue |
E3S Web Conf.
Volume 582, 2024
1st International Conference on Materials Sciences and Mechatronics for Sustainable Energy and the Environment (MSMS2E 2024)
|
|
---|---|---|
Article Number | 03002 | |
Number of page(s) | 13 | |
Section | Mechatronics in Energy | |
DOI | https://doi.org/10.1051/e3sconf/202458203002 | |
Published online | 22 October 2024 |
Metamodeling for predicting the behavior of airfoils of wind turbine blades: An integration of artificial neural networks
1 Laboratory of Innovative Technologies, National School of Applied Sciences of Tangier. B.P. 1818, Tangier, Morocco.
2 Team of engineering and applied physics, Higher School of Technology of Beni Mellal, Morocco.
3 Unité de Formation et de Recherche Laboratoire de Mécanique et informatique, Université Félix Houphouët-Boigny, Abidjan, Cote D’Ivoire.
* Corresponding author: haiek.mohammed@gmail.com
The aim of this paper is to develop a robust metamodel capable of predicting the behavior of wind turbine blade airfoil profiles, despite variations in the input parameters of computational fluid dynamics (CFD) simulations. The metamodel development begins with the specification of essential geometric parameters for the simulations. Subsequently, an empirical analysis of the airfoil profiles is conducted, and the results of the corresponding simulations are presented. These data are used to train and refine the metamodel, which is based on an artificial neural network. The model fitting process is divided into three main stages: training, validation, and testing, during which we strive to minimize the error function using the Levenberg-Marquardt algorithm. In conclusion, we validate our model by a thorough comparison of the results from the metamodel and the CFD simulations, aiming to optimize computation time.
© The Authors, published by EDP Sciences, 2024
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.