Marshall–Olkin Power Lomax distribution for modeling of wind speed data

Abstract

Accurate collection of wind speed records is significant for numerous wind power applications. The present investigation aims to highlight the use of the Marshall–Olkin Power Lomax (MOPLx) distribution for wind speed data. We examine the actual wind speed records gathered from three stations Bahawalpur, Gwadar, and Haripur. The dataset is demonstrated by using MOPLx distribution and compare its modeling performance with renowned probability distributions, for example, Weibull–Lomax, power Lomax, Weibull, power Lindley, Lindley, and Lomax. Findings indicate that MOPLx distribution gives the best fitting as per model evaluation criteria, Akaike information criterion (AIC), Bayesian information criterion (BIC), Kolmogorov Smirnov test (KS), coefficient of determination (R2) and root mean square error (RMSE). Overall, the results demonstrate the feasibility, precision, and effectiveness of the MOPLx distribution for portraying wind speed modeling. It is also observed that the MOPLx model is ideal in terms of the power density error (PDE) criterion.