Inverted Kumarswamy distribution for modeling the wind speed data: Lake Van, Turkey

Abstract

In calculating the wind energy potential of a region, some important points such as determining the distribution used to model wind speeds and estimating the parameters of the distribution accurately should be considered. Many different distributions have been proposed in wind energy literature over the years. In this paper, some of these studies are reviewed. Then, Inverted Kumaraswamy (IKum) distribution is used for the first time to model wind speed data as an alternative to the well-accepted Weibull distribution. Maximum Likelihood, Least Squares, and Maximum Product of Spacing methodologies are employed in estimating the parameters of the IKum distribution. A Monte Carlo simulation study is conducted for comparing the efficiencies of these methods. The wind speed data sets considered in this study include wind speeds from 6 stations located around Lake Van in Turkey. Modeling performances of the Weibull and IKum distributions are evaluated with the well-known goodness-of-fit criteria and power density error values. Results show that the IKum distribution can be considered as an alternative to the well-accepted Weibull distribution.