Forecasting solar radiation during dust storms using deep learning.

TLDR: In this paper, the authors deal with the analysis of solar radiation and power output of a rooftop photovoltaic plant during a dust storm and propose a forecasting methodology using deep learning network.

Abstract: Dust storms are common in arid zones on the earth and others planets such as Mars. The impact of dust storms on solar radiation has significant implications for solar power plants and autonomous vehicles powered by solar panels. This paper deals with the analysis of solar radiation and power output of a rooftop photovoltaic plant during a dust storm and proposes a forecasting methodology using deep learning network. The increased aerosol content due to dust storms increases the diffuse component of the solar radiation. This effect persists for a long duration and can impact the quality of forecasting of solar radiation. Deep learning networks that capture long range structure can improve the quality of solar radiation forecasting during dust storms. These results can help explain the sudden drop in power output of solar plants due to dust storms originating in another continent. They can shed light on mysterious cleaning events in autonomous vehicles powered by solar panels to be used in space missions.