Eventos Anais de eventos
ENCIT 2018
Brazilian Congress of Thermal Sciences and Engineering
DETERMINATION OF WEIBULL CURVE PARAMETERS APPLIED TO WIND ENERGY BY PARTICLE SWARM OPTIMIZATION
Submission Author:
MARCUS VINICIUS SILVEIRA MACEDO , CE , Brazil
Co-Authors:
MARCUS VINICIUS SILVEIRA MACEDO, LINDEMBERG FERREIRA DOS SANTOS, Felipe Ferreira Gomes, Carla Freitas de Andrade, Paulo Alexandre Costa Rocha
Presenter: MARCUS VINICIUS SILVEIRA MACEDO
doi://10.26678/ABCM.ENCIT2018.CIT18-0735
Abstract
Wind power is one of the most widely used renewable energy sources in Brazil. To evaluate the wind potential of a region, a wind behavior model is required. Several probability distributions are used to represent the wind speed data, such as Weibull, Gamma and Normal, and Weibull is the most used for this purpose. This distribution is dependent of two parameters, shape factor k and scale factor c, usually estimated by deterministic methods. Studies have shown that some deterministic methods lead to an unsatisfactory fit. This paper aims to estimate the Weibul distribution curve parameters for the region of Triunfo-PE, by the heuristic method known as Particle Swarm Optimization Method. The method was implemented by R language and the results were compared using the statistical tests RMSE, MAE and R2, as well as the percentage value of the production deviation between the curve obtained and the histogram generated from the wind speed data - WPD. Particle Swarm Optimization method did not prove to be efficient to estimate the Weibull curve parameters for the analyzed region, since it obtained high values of RMSE, MAE and low value of R2 when compared to the deterministic methods. The WPD equal to 10.88% was higher than the acceptable maximum for the deviation of production.
Keywords
wind energy, Weibull Distribution, Particle Swarm Optimization, Heuristic, Deterministic
