Eventos Anais de eventos
ENCIT 2022
19th Brazilian Congress of Thermal Sciences and Engineering
Sensitivity analysis of small-scale vertical-axis wind turbines: improving the design space using Bézier-GAN
Submission Author:
Gabriel Bertacco dos Santos , SP , Brazil
Co-Authors:
Gabriel Bertacco dos Santos, Leandro Salviano
Presenter: Gabriel Bertacco dos Santos
doi://10.26678/ABCM.ENCIT2022.CIT22-0136
Abstract
Wind energy has emerged as an attractive alternative to the current fossil-fuel-based energy mix. The rapid growth of wind power plants around the world has inspired the development of novel technologies and more efficient equipment to meet the world’s ever-growing energy needs. In this context, small-scale H-Darries vertical-axis wind turbines (VAWTs) are most appealing for harvesting wind energy in urban-like conditions. Nonetheless, the overall efficiency of H-Darries VAWTs still struggles with different complex aerodynamic phenomena. Seeking to improve the way optimization procedures for H-Darries VAWTs are currently done, we investigate the application of the BézierGAN parameterization in this context. For that, we adopt 3D computational fluid dynamics (CFD) simulations coupled with two sensitivity analysis methods. The results show that the BézierGAN solves the trade-off between shape variability and number of input parameters, providing high shape variability in a bounded, non-interactive, low-dimensional latent space. Therefore, we recommend this recently proposed parameterization method for future research and optimization of H-Darries VAWTs.
Keywords
renewable energy, vertical axis wind turbine, generative neural network, airfoil parameterization, Sensitivity analysis
