Estimation of Weibull Distribution Parameters by Analytical Methods for the Wind Speed of Jhimpir, Pakistan - A Comparative Assessment

Abstract

Assessing the potential of a wind farm requires looking into how the wind behaves throughout a certain time frame. One of the most popular ways to statistically model wind data is with the Weibull distribution. Estimating two parameters of the Weibull PDF is crucial for a better fit between the PDF and wind speed data. In this study, Weibull distribution parameters for 2019 wind speed data in the Jhimpir region of Pakistan are determined using four analytical techniques: the empirical method (EM), the maximum likelihood method (MLM), the method of moments (MoM), and the energy pattern factor (EPF) approach. Each technique is evaluated using several different metrics, including the root mean squared error (RMSE), mean absolute error (MAE), mean absolute relative error (MARE), and the coefficient of correlation (R). Statistical analyses show that the shape (k) and scale (c) parameters of the Weibull distribution estimated by the EM, MLM, and MoM are quite close to one another compared to the ones obtained by EPF for the available data. The MATLAB environment-based numerical results expressed that the EPF method performed the best in terms of R and RMSE and worst in terms of MAE and MARE.