Wind speed estimation by effective parameters using different regression models

Wind energy is the most preferred renewable energy source by developing countries due to the fact that the installation of wind power plants is fast and the costs are low and they do not have negative effects on the natural environment. The most important input in wind energy conversion systems is wind speed. Therefore, it is necessary to determine the wind speed accurately and reliably in order to obtain maximum efficiency from the wind energy. Wind speed is a difficult parameter to predict as it has many environmental factors affecting its change. For this reason, meteorological data collected through established measuring stations are used for wind speed estimation. The data used in this study was obtained from the measurement station established in the campus of the Faculty of Engineering and Natural Sciences of Tokat Gaziosmanpaşa University. The wind speed, humidity, pressure and temperature information measured by the sensors in the measuring station at intervals of ten minute are grouped monthly, seasonal and annual to be used for analysis. In this study, it is aimed to estimation by making a mathematical connection between meteorological data and wind speed. For this purpose, monthly, seasonal and annual equation sets were formed by the analysis. While determining the equation sets, wind speed estimation is made with six different combinations of temperature, pressure and humidity parameters. Simple Linear Regression, Multiple Linear Regression and Multiple Non-linear Regression methods were used in the analyzes for estimation. When the root of the mean squared error and the coefficient of determination were taken into consideration, it was observed that multiple non-linear regression methods were prominent and estimated with lower error. It was determined that the temperature and humidity parameters were used together as an input for best wind speed estimation.