Multi-Day-Ahead PV Power Output for Improved Radial Basis Function Neural Network

Photovoltaic (PV) power output forecasting is a crucial issue in the PV power industry and power company. Nevertheless, the disadvantage of the existing PV power output forecasting models is that they often ignore weather fluctuation and uncertain information between the irradiation and ambient temperature caused by moving cloud, which leads to short of forecasting accuracy. This paper is to present an improved radial basis function neural network structure (RBFNN) with multi- parameters forecasting scheme to overcome above drawbacks. Result of the case study acquired by the proposed scheme is compared with two typical artificial neural network forecasting methods including radial basis function neural network (RBFNN) and back propagation neural network (BPNN). It demonstrations that the proposed model leads to better precision for forecasting PV power output.