Forecasting the AC power output of a PV plant accurately is important both for plant owners and electric system operators. Two main categories of PV modeling are available: the parametric and the nonparametric. In this paper, a methodology using a nonparametric PV model is proposed, using as inputs several forecasts of meteorological variables from a Numerical Weather Forecast model, and actual AC power measurements of PV plants. The methodology was built upon the R environment and uses Quantile Regression Forests as machine learning tool to forecast AC power with a confidence interval. Real data from five PV plants was used to validate the methodology, and results show that daily production is predicted with an absolute cvMBE lower than 1.3%.
Phenomena of overirradiance have been pointed all over the World. This note presents the most extreme enhancement event reported in Brazil, which contains an irradiance reading of 1590 W/m 2 measured in Sao Paulo (latitude 23°32'S) at relatively low altitude (760 m a.s.L).
Forecast procedures for large ground mounted PV plants or smaller BIPV or BAPV systems may use a parametric or a nonparametric model of the PV system. In this paper, both approaches are used independently to calculate the energy delivered to the grid on an hourly basis in forecast procedures that use meteorological variables from a Numerical Weather Prediction model as inputs, and their performances against real generation data from six PV plants are analyzed. The parametric approach relies on mathematical models with several parameters that describe the PV systems and it was implemented in MATLAB®, whereas the nonparametric approach is based on Quantile Regression Forests with training and forecast stages and its code was built in R. The parametric approach presented more significant bias on its results, mostly due to the input data and the transposition model of irradiance from a horizontal surface to the plane of the PV array.
Acima de tudo agradeço a Deus, pelas oportunidades de estudo, trabalho e crescimento como pessoa. Aos meus pais, Jorge e Fátima, que, mesmo distantes, me dispensaram muito amor, carinho e incentivo. A minha irmã Flávia, pelas visitas frequentes. Aos meus tios João, Zulmira e Nancy, pelo apoio dado. Ao meu orientador, Prof. Zilles, pelos conhecimentos e experiências repassados, pelas oportunidades concedidas e pela dedicação dispensada. A minha namorada, Gabi, pelo carinho e paciência. Aos meus companheiros de LSF, por toda a ajuda nesses anos de mestrado. Aos meus amigos, pelos momentos de felicidade que passamos. À Universidade de São Paulo, pelo bom ambiente de estudo proporcionado. Ao INCT EREEA, pelo apoio financeiro através do CNPq e da CAPES.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.