出版社：Grupo de Pesquisa Metodologias em Ensino e Aprendizagem em Ciências
摘要：The search for renewable energy has increased in recent decades, mainly due to environmental concerns as well as the growing energy demand. Wind energy is one of the sources of renewable energy widely adopted in the world, with the northeast region of Brazil highlighted by its large production capacity. The implementation of wind farms lacks preliminary studies on statistical modeling of wind speed and, among the proposed methods, the Markov Chain is known to be efficient. This work presents a statistical analysis of the wind speed in the city of Petrolina - PE, with the objective to apply a first-order Markov chain, a probabilistic method, in the hourly wind speed modeling. Data were obtained from the website of the National Institute of Meteorology (INMET), corresponding to the period 2010 to 2020. First-order Markov chain with two states was used to find transition probabilities, and anual variation of these probabilities and normalized Entropy was analyzed. The results showed that the proposed model adjusts well to wind speed variations in specific months of the year and that models with more states can be proposed to reduce the loss of information calculated by the entropy value.
