Simulation of the Wind Speed at Different Heights Using Artificial Neural Networks

Abstract

In this paper, an estimation of the wind speed at different heights with artificial neural networks is presented. It is an alternative way to compute wind shear. This method was tested using pairs of data sets from two measuring stations, installed in different topographic locations. Wind speed simulation is performed with high accuracy. The calculation of the surface friction coefficient from the actual measurements is also compared for wind shear estimation with the typical method in terms of energy output. Results showed that artificial neural networks achieve a better wind speed simulation and wind power estimation at different heights, even in complex terrains.