Parameter estimation of weibull probability distribution by seven methods – a wind regime of the city of nitra, Slovakia

Slovakia currently has a relatively large unused potential in the area of electricity production from solar radiation and wind as renewable sources. The conversion of the wind’s mechanical energy into electrical energy depends, among other things, on the wind speed and its turbulence. Perhaps the most widely used probability distribution for a wind speed model is the Weibull distribution. In the article, we deal with the comparison of seven methods for estimating the parameters of this distribution – maximum likelihood method, method of moments, empirical method, empirical method of Lysen, power density method, least squares method and weighted least squares method – on wind speed records from the city of Nitra for the period of 2005-2021. The vicinity of this city is one of the places identified as a suitable location for the installation of wind turbines. The performance of individual estimation methods is evaluated based on the indicators – the coefficient of determination R 2 and the root mean square error RMSE. Based on these values, the most accurate method is the weighted least squares method, although all other methods achieved similarly good results.

---

Opracowanie rekordu ze środków MNiSW, umowa nr POPUL/SP/0154/2024/02 w ramach programu "Społeczna odpowiedzialność nauki II" - moduł: Popularyzacja nauki (2025).