In the literature, it is known that Weibull is the most
widely-used distribution to estimate wind energy potential
because of its flexibility and easy computation [1–13]. However,
it should be noted that the Weibull distribution is unable to model
all the wind structures encountered in nature. For this reason, in
recent years, in order to model wind speed data more smoothly,
the use of a variety of statistical distributions has been proposed
in a large number of studies. For instance, [14] considers
normal-mixture distribution and mixture Weibull distribution versus
the classical Weibull distribution. [15–17,14,18] introduce two
mixture Weibull distribution to model wind speed data. [19–22]
propose the distributions derived from the maximum entropy
principle. Similarly, [23] and [24] respectively introduce distributions
derived from the maximum entropy principle and minimum
cross entropy principle. [25] presents a comparison of log-normal,
gamma, Weibull and Rayleigh models. [26] provides a review and
literature on the usage of various statistical distributions in modeling
wind speed data. As well as the mentioned statistical distributions,
Erlang, inverse normal and gumbel-maximum distributions
are presented as wind speed distributions in [27], while a generalized
extreme value distribution is used in [28]. On the other hand,
[29] firstly introduces mixture Gamma–Weibull and mixture truncated
normal distributions, while [30] proposes certain flexible
families of distributions as an alternative to the Weibull