Performance analysis of Weibull methods for estimation of wind speed distributions in the adamaoua region of Cameroon

Abstract

This paper explores the performance analysis of five Weibull distribution methods to select the more accurate estimation for the Weibull parameters using time-series of measured daily wind speed data collected in three localities in the Adamaoua region of Cameroon. The Weibull distribution with two parameters, the shape k, and the scale C, was specifically considered to be a good quality probabilistic model for wind speed distributions. The five Weibull distribution methods proved to be effective in evaluating the parameters of the Weibull distribution. This fact was supported by the values of the root mean square error, the Chi-square (χ²) and the correlation coefficient R² which showed magnitudes very close to each other. In addition, the comparison between wind speed distributions predicted by the Weibull methods and wind speed distributions measured locally, suggested that the most accurate two-parameter Weibull distribution method is the Energy Pattern Factor Method (EPF). As a result, to reduce uncertainties related to the wind energy output calculation, the EPF is recommended for estimating wind speed distributions.