Uso de séries temporais para previsão de irradiação solar

Abstract

Efficientplanninganddecision-makingprocessesrelatedtophotovoltaicsolarpowerplants increasingly demand an accurate knowledge of solar irradiance. The analysis of time series applied to climatological data has attracted special interest in recent years, since the climate directly interferes in many economic activities. This work aims to develop and compare models of SARIMA time series to forecast solar irradiance, based on data from a meteorological station located in Maceió-AL. Based on the irradiance data, the Box-Jenkins forecast method was applied to identify a mathematical model. To estimate the model parameters, the maximum likelihood method was used. To select the most appropriate models, well-defined metrics were used,such as the mean square error (MSE). Hence,it was possible to make the forecast for the day ahead of the studied time series and evaluate it with error analysis methods. Finally, in this work it was possible to describe and implement prediction techniques in the time series with collected data, enabling a real case study, where it was possible to conclude that the models built are satisfactory for forecasting a day ahead from the established metrics and limits