WEIBULL PARAMETERS ESTIMATION FOR WIND SPEEDPROBABILITY DISTRIBUTION IN KON DONG USING FIVE DIFFERENT NUMERICAL METHODS

Abstract

Wind energy, a clean and sustainable renewable resource, plays a crucial role in advancing global sustainable development. Accurate wind speed estimation
is essential for assessing the power output potential of wind turbines. This study models wind speed data from the Kon Dong wind farm using statistical methods
based on the Weibull distribution. Five methods for estimating the Weibull shape and scale parameters were evaluated: the Energy Pattern Factor Method
(EPFM), Empirical Methods by Lysen (EML) and Justus (EMJ), a hybrid EPFM-EMJ approach, and Method of Moments (MoM). Model performance was assessed
using root mean square error (RMSE) under both non-seasonal and seasonal conditions. Results show that EPFM, EMJ, EPFM-EMJ, and MoM achieved good fits,
with RMSE values ranging from 0.022 to 0.024 for non-seasonal data, 0.024 to 0.026 during the strong wind season, 0.005 in the rainy season, and 0.020 to
0.023 in the transitional season. In contrast, the EML method consistently produced the highest RMSE across all conditions, indicating the poorest fit. These
findings highlight the effectiveness of EPFM, EMJ, EPFM-EMJ, and MoM in accurately estimating Weibull parameters for wind resource assessment.