Estimation

Abstract

The penetration of non-conventional renewable energies for electricity generation, especially solar power plants that use photovoltaic cells, has experienced considerable growth in Latin America and the Caribbean since the beginning of the 21^st century. It is gaining importance in electricity matrix of various countries in the region. The integration of these power plants poses several critical challenges that affect the grid stability and electricity dispatch management due to the intermittency of the main resource used by these plants which is the surface solar irradiance. To address these challenges, various power estimation and/or SSI estimation models based on neural networks such as recurrent neural networks and convolutional neural networks, statistical methods, and hybrid algorithms have been developed. The performance of these models depends intrinsically on the nature of the data input, which includes satellite images, meteorological variables, and power time series. This article presents a review of the current models presented in previous publication, establishing a comparative framework on the origin of the data in the development of solar estimation models.