Forecast of solar radiation with the application of neural networks in rural zones of Ecuador

Currently, advances in renewable energy in systems connected to the distribution network or in the form of an island have become an alternative to generate electricity in rural sectors in developing countries, so this research uses the data of measurements 3 years of an Ecuadorian rural area of the community of Saraguro-Uchucay, to predict solar radiation through the use of data analytics and artificial intelligence to determine the photovoltaic solar generation capacity of the study sector. The models are made on the WEKA software platform through the analysis of a matrix of variables that seeks the best algorithm to determine the incidence of daily solar radiation and establish a model of electricity production to meet the demand for electrical energy.