A novel deterministic model for generating synthetic wind speeds

This work proposes a deterministic methodology to generate synthetic wind-speed time series consistent with both a probability density function and a spectral density function. The method relies on the random-phase multisine signal to generate an initial sequence conforming to the target spectral density. A following iterative rank-reordering procedure rearranges the samples drawn from the target probability distribution so as to match the desired spectral density function. Such a rank-reordering procedure generates a final signal that conforms to a stationary, pseudo-random process. An application of the method is presented along with a comparison with two different state-of-the-art algorithms from the literature.