A fast and effective procedure for sensing solar irradiance in photovoltaic arrays

Photovoltaic cells are characterized via a static relationship that describes their current-voltage dependency. This relationship depends on two critical uncertain parameters: temperature and solar irradiance. While it is technologically feasible and easy to measure the former, a sensor for the latter is usually expensive and difficult to calibrate. In this work we show that estimation of solar irradiance can be accomplished in closed form, assuming that current, voltage and cell temperature are available for measurement. The approach is validated and compared with a neural network based device, by using outdoor measurements taken in a partially overcast day with rapidly moving clouds, demonstrating the ability of the estimator to dynamically track irradiance.