Short-term PV power forecasting using Support Vector Regression and local monitoring data

Abstract

In recent years many research works have study the problem of photovoltaic power forecasting because of its importance to grid management and large-scale PV integration. In order to forecast the Photovoltaic power production in the region of Casablanca Morocco, a simple and reliable model based on Support Vector Regression (SVR) and local monitoring data is proposed in this paper. Three models based on ε-SVR, ν-SVR and LS-SVR are compared using five performance indicators, MAE, MSE, RMSE, R2 and RRMSE (%). The best model shows a good results with an RRMSE of 15.23% and a coefficient of determination R2 = 0.96%.