Title: On the Performance of Forecasting Models in the Presence of Input Uncertainty

Abstract: Nowadays, with the unprecedented penetration of renewable distributed energy resources (DERs), the necessity of an efficient energy forecasting model is more demanding than before. Generally, forecasting models are trained using observed weather data while the trained models are applied for energy forecasting using forecasted weather data. In this study, the performance of several commonly used forecasting methods in the presence of weather predictors with uncertainty is assessed and compared. Accordingly, both observed and forecasted weather data are collected, then the influential predictors for solar PV generation forecasting model are selected using several measures. Using observed and forecasted weather data, an analysis on the uncertainty of weather variables is represented by MAE and bootstrapping. The energy forecasting model is trained using observed weather data, and finally, the performance of several commonly used forecasting methods in solar energy forecasting is simulated and compared for a real case study.