Statistical Models to Predict Solar Radiation At High Resolutions
Abstract
In this work, we assess the performance of statistical models for intra-hour solar forecasting. More precisely, a linear recursive model and a nonlinear model are used for generating solar irradiance forecasts at temporal resolutions of a few minutes and over multiple horizons. Our approach is applied to forecasting solar irradiance at single sites using the sole historical ground observations of solar irradiance. The benchmarking of the forecasting methods is made at four sites that exhibit different sky conditions.
Keywords
Intra-hour solar forecasting, Statistical models