Interval Power Forecasting for Photovoltaic Charging Station in Renewable Power System

Load forecasting of photovoltaic charging stations in renewable power system has grown in importance as the global energy structure has changed and new energy vehicles have gained popularity. Current forecasting methods still face challenges such as data quality, model generalization capability and accuracy. In this research, we offer a Multilayer Extreme Learning Machine approach for predicting the interval power of photovoltaic charging stations that takes into account both weather and charging user demand response. The sparrow optimization algorithm is used to optimize the parameters of the model, and quantile regression is used as the loss function of the prediction model to achieve the load interval prediction of PV charging stations under different confidence levels. The prediction approach suggested in this study excels in interval prediction accuracy, according to the findings of the algorithm assessments.