A high concentrated photovoltaic output power predictive model based on Fuzzy Clustering and RBF neural network

Abstract

It is significant to make a short-term power output forecast for the solar photovoltaic power station. On the one hand it helps guarantee power grid security, on the other hand it can increase the efficiency of power generation. This paper designs a high concentrated photovoltaic output power prediction model based on the Fuzzy Clustering and Radial Basis Function (RBF) neural network, uses the meteorological data which affect output power to classify the sample collection, selects the most similar two days' data and daily current average radiation exposure as the RBF neural network inputs. The output value of the neural network is the unit's prediction output power after an hour. This paper uses the data from a high concentrated photovoltaic power station in northwest China to train and validate the model, the predicting outcomes show the proposed model has good accuracy.