Exploitation of a New Short-Term Multimodel Photovoltaic Power Forecasting Method in the Very Short-Term Horizon to Derive a Multi-Time Scale Forecasting System

The relentless spread of photovoltaic production drives searches smart approaches to mitigate unbalances in power demand and supply, instability on the grid ensuring stable revenues producer. Because development energy markets with multiple time sessions, there is a growing need forecasting for steps, from fifteen minutes up days ahead. To address this issue, study both short-term-horizon three very-short-term-horizon hours methods are presented. short-term based multimodel approach referred several configurations Analog Ensemble method, using weather forecast four numerical prediction models. very-short-term consists an Auto-Regressive Integrated Moving Average Model eXogenous input (ARIMAX) that uses irradiance satellite elaborations as exogenous variables. methods, applied one year small-scale grid-connected plants Italy, have obtained promising improvements respect refence methods. horizon after which was able outperform has also been analyzed. revealed usefulness data cloudiness properly interpret results performance analysis.