Global solar radiation forecast using an ensemble learning approach

<span lang="EN-US">With the increase in demand for solar power, a power forecasting model is of maximum importance to allow higher level integration non-conventional energy into existing electricity grid. With advancement data availability, there’s good time use data-driven algorithms enhanced prediction generation. Gathering and analyzing can predict generation mitigate impact intermittency. During this research, we explore automatically creating models that are site-specific utilizing machine learning generate radiation from meteorological station weather forecast reports, predicted corresponding output be calculated depending upon characteristics PV system used. The challenge enhance accuracy forecast. Ensemble techniques like random forest (RF) extreme gradient boosting (XGBoost) well suited as they improve stability combine several reduce variation bias which outperforms majority models, result making them perfect field prediction.</span>