A prediction method for photovoltaic power generation with advanced Radial Basis Function Network

This paper proposes a new method for short-time generation output prediction of PV systems. The proposed method is based on the hybrid intelligent system of GRBFN (Generalized Radial Basis Function Network) of ANN (Artificial Neural Network) and DA (Deterministic Annealing) Clustering. RBFN is one of ANNs that provide the good performance. However, it has an open problem in constructing RBFN with the good accuracy. To improve the performance, this paper introduces two strategies: one is to use DA of global clustering to select the good initial values of the center and the width of radial basis functions and the other is to use GRBFN to determine the center and the width through the learning process appropriately. As a result, the proposed method provides better results than the conventional ones. The proposed method is successfully applied to real data of short time prediction of PV systems.