Example for

Weibull fitting

Weibull fitting

WindDataSuite implements a module for fitting
binned wind speed frequencies to
Weibull distributions.
The fitting can be calculated for the 2-parameter Weibull distribution
(scale parameter A, shape parameter k) and for the 3-parameter
Weibull distribution (scale parameter A, shape parameter k,
location parameter v_{0}).

Besides the methods of the linear regression and the Maximum Likelihood model, the Weibull distribution can also be determined via the best possible fit to the mean value.

In the following example, the measured wind speed data have been reduced to complete profiles up to 150 m height with the WindDataSuite module for data filtering and have then been binned with respect to the wind speed frequencies. Then, the resulting binned frequency distributions have been fitted with the Maximum-Likelihood model to the 2-parameter Weibull distribution.

Fig.1:

Frequencies of the wind speeds

Fig.2:

Frequencies of the wind speeds of the fitted Weibull distributions

Fig.3:

Measured and Weibull fitted wind speeds

Fig.4:

Shape parameter k of the Weibull distributions

1 year (365 days), sampling rate 10 minutes

21 height levels of wind speed, complete profiles only

starting from the time series approx. 3 seconds

starting from the time series approx. 2 minutes