Maximum wind speed forecasting using historical data and artificial neural networks modeling

Abstract

Estimation of the wind speed makes a very important contribution to the seamless integration of wind power plants into the grid. In this way, the maximum amount of electricity can be generated by estimating the amount of energy that can be generated from wind energy. The measurements of the wind speed in the region, where the plant is plant to be established, made before the installation of the wind power plants (WPP), takes between 6 and 18 months. In this study, it was investigated what could be done to make a foresight and estimation about the wind speed in the future for the selected region. In order to accurately determine the wind speed, it was tried to be estimated by using artificial neural networks (ANN) included in the MATLAB package program. In this study, 365 data belonging to the previous years of the region to be studied were provided and used to train the ANN of the planned study. In practice, the parameters of temperature, humidity and pressure, which are among the factors affecting wind speed, were taken into consideration. An R value of 91.20% in the training phase, 93.04% in the validation phase and 92.76% in the test phase was obtained. High accuracy values were obtained at all phases and it was shown in this study that ANN can estimate reliably without memorizing.