Bayesian analysis of geo-dependencies in wind speed

Abstract

The estimation of the wind resource had a large increase in recent years due to the strong demand for renewable energy. Wind estimation is a dynamic problem due to its stochastic nature. Consequently, the study of wind speed in different places helps to know the geospatial dependencies that exist between them. In this paper, we present a stochastic process that analyzes the geographic dependencies in eight sites in the Metropolitan Area of Queretaro using a statistical approach using the Expectation-Maximization (EM) algorithm. The Spearman correlation coefficient is then used to determine the intensity in the relationship between each of the sites. The results of the experimentation were carried out in three scenarios, in different periods, in this way we can know the type of association in the sites during each period. In this way, the sites where wind power can be used to obtain wind energy are known.