Global Horizontal Irradiance Prediction Model Based on Mixed Spatial Information and Aerosol Classification

Reliable and accurate predictions of solar radiation are essential for the supervision and operation of solar photovoltaic power generation systems. As the primary media involved in atmospheric radiation transfer, aerosols significantly influence global horizontal irradiance (GHI). The composition, shape, and number density distribution of aerosols vary greatly, resulting in significant differences in their optical properties, which in turn affect solar radiation in different ways. This study aims to explore the impact of different types of aerosols on predicting GHI. First, we expanded the data within a fixed region by incorporating spatial information to supplement the timescale data. Furthermore, we used the Informer model to forecast the GHI in different regions, inputting historical data on aerosol optical depth (AOD), meteorological parameters, and GHI. Finally, we used an aerosol classification model to classify aerosols in different regions and calculated the GHI predictions for different aerosol types. The findings suggest that aerosol classification impacts the predictive performance of the GHI. When continental and subcontinental aerosols dominated, the predictive performance of the GHI improved. When biomass-burning aerosols dominate, the predictive accuracy of the GHI reduced.