This paper demonstrates a design method that integrates various computational tools such as Rhinoceros, the artificial neural network from MATLAB, and computational fluid dynamics from Eddy3D to search for the optimal aerodynamic geometries for a wind turbine. It introduces a site-specific microclimate analysis method that can maximize site-specific wind energy potential. Through the integration of computational fluid dynamics (CFD) and artificial neural networks (ANN), the study was able to find the optimized shape to maximize the wind potential for the specific test site. These ANN models use fewer computational resources and less time with reasonable average regression values up to 0.96. The result shows improvement of the annual hourly wind speed around the wind turbine up to 13.24 m/s. It would be beneficial to test the proposed method with actual performance to improve the proposed method.
