Short-term photovoltaic output forecasting model for economic dispatch of power system incorporating large-scale photovoltaic plant

A combined prediction method based on ensemble empirical mode decomposition (EEMD) and support vector machine (SVM) is proposed to tackle with the problem of the short-term forecast of photovoltaic system (PVs) hourly output a day ahead. Weather types are divided into abnormal day (weather changed suddenly) and normal day. By the proposed method, firstly, the history data for hourly output of PVs is decomposed into a series of components by using EEMD method. Considering different factors for different type of weather, different models are built and different kernel functions and parameters are chosen to deal with each component of the decomposed data by using SVM. Simulation results show that the proposed classification modeling ideas and EEMD-SVM combination forecasting method enable that the mean absolute percentage error results for the abnormal days is decreased by 5%, and for normal day is decreased by 3% comparing with the traditional SVM method and Back Propagation (BP) neural network method respectively.