Voltage Prediction of Substation Users with Photovoltaic Based on Weather Classification

Due to the significant impact of meteorological factors on photovoltaic users, it is necessary to reasonably consider the influence of meteorological factors in voltage prediction in substation users with photovoltaic. This article first analyzes the correlation between meteorological factors and photovoltaic users voltage. Then considers that different meteorological factors have different degrees of correlation with voltage, principal component analysis is used to obtain coupled meteorological factors, cluster the coupled meteorological factors, and finally obtain historical voltage data under different meteorological conditions. The Gated Recurrent Unit (GRU) model is selected to predict the user voltage, considering the current weather type. The prediction results indicate that the model using weather classification can provide more accurate voltage prediction and has good practicality.