<b>Adjustment of models of probability distribution to hourly wind speed series for Ponta Grossa, Paraná State</b> - doi: 10.4025/actascitechnol.v33i4.7072

Abstract

The objective of this work was to evaluate the adjustment of data series of hourly wind speed for Ponta Grossa, Paraná State, to the probability distributions of Weibull, Rayleigh, Beta, Gama and Normal and to analyze possible interferences of the day and night atmospheric conditions on the adjustment of these models. The test of Kolmogorov-Smirnov was used to compare the adjustments and to select the best theoretical distributions inside of each one of the twenty-four hours of the day in every month of the year, regarding the period of January of 1998 to December of 2007. The monthly hourly asymmetry coefficients were also used to characterize the distributions. The largest or smaller degree of adjustment of the hourly wind speed data to the appraised models were influenced by the hour of the day, and, in the night period, between zero and five hours, the distribution of Weibull presented the best adjustment, followed by Beta and Gama distributions. In the period of the day, the five appraised models adjusted the hourly wind speed data appropriately, having however, alternation for the order of the best adjustment.