SILVA, L. P.; http://lattes.cnpq.br/5852568421266103; SILVA, Leide Pricila da.
Résumé:
In this research, the parameters of shape and scale of the Weibull distribution are estimated,
using the Graphical (MG), Empirical (ME), Moments (MM), Maximum Likelihood (MLE) and
Power Density (MDP) methods. For that, anemometric data from the SONDA Network
(National Organization System for Environmental Data), from Petrolina-PE, São João do
Cariri-PB, Triunfo-PE, Natal-RN, Belo Jardim-PE and São Luiz stations were used -MA,
Brazil. These data are obtained at heights of 25 and 50 meters, with a measurement interval of
5 seconds, under which averages were obtained every 10 minutes. Environmental data obtained
at 10 m were also considered, also with an interval of 5 seconds and extraction of averages
every minute. In a second step, estimates of power and energy density were made. To examine
whether the theoretical distribution function is adequate to describe the wind speed data, the
Mean Absolute Error (EAM), the Mean Relative Error (ERM) and the Root Mean Square Error
(REQM) were adopted. It was found that the methods used in the research showed good
agreement with the data measured at both times. In Petrolina, the Average Power Density-SD
and the Energy Density-DE at 10 m showed average values of 20.41 Wm-2 and 53.65 MJm-2
respectively. And, at 25 and 50 meters, DP and DE equal to 42.43 Wm-2, 121.74 MJm-2 and
90.71 Wm-2 and 260.36 MJm-2, respectively, were obtained. The Annual Average Power
Density (DPMA) in Natal-RN was 70,885 Wm-2 (186,176 MJm-2) in 2010 and 63,077 Wm-2
(165.7 MJm-2) in 2011, in São Luiz-MA, 2016, 28,002 Wm-2 (73,652 MJm-2). In the city of SJ
do Cariri-PB, the values of DPMA (DEMA) at 25 m for the years 2006, 2007 and 2008 were,respectively, 89.844 Wm-2 (236.429 MJm-2), 109.611 Wm-2 288.550 (MJm-2) and 91.478 Wm-
2 (240.298 MJm-2) already at 50 meters for the same years, 130.108 Wm--2 (342.398 MJm-2),
152.577 Wm-2 (401.654 MJm-2) and 129.768 Wm-2 (340.911 MJm-2). For Belo Jardim-PE,
2004, at 25 m and 50 m the estimated values of DPMA and DEMA were, respectively, 95.339
Wm-2 (250.069 MJm-2) and 103.330 Wm-2 (270.948 MJm-2). Triunfo was the station that
presented the highest value of the average annual speed (13.654 ms-1) and thus, also, the highest
DPMA 1899.450 Wm-2 (4991.466 MJm-2). The results indicate that the best performance was
obtained by MLE, followed by MM, at both times, although the metrics used do not express a
high difference between all the tested methods. The biggest errors were presented by MG,
followed by MDP, for data at 10 m and by MDP followed by MG at 25 m and 50 m. Therefore,
considering the objectivity and better performance that the method presents, the use of MLE is
indicated.