Wind resource analysis and characterization with Markov’s-transition matrices

Wind resource analysis and characterization is a key ingredient in the planning and operation of wind farms. The most popular method to perform it is based on the generation of histograms showing the frequency of occurrence of the wind speed and their approximation with probability density functions (PDF). This paper presents the mathematical bases and examples of application of two statistical methods using a matrix based representation. The main advantage of this representation is that allows the preservation of information related to the temporal patterns of the wind speed time series and the results are shown. Further studies are being carried out to use these matrixes to reconstruct and forecast the wind speed.