Study on Control Strategy for Photovoltaic Energy Systems Based on Recurrent Fuzzy Neural Networks

To increase the output efficiency of a photovoltaic (PV) energy system, the real-time maximum power point (MPP) of the PV array must be tracked closely. Herein a recurrent fuzzy neural network controller (RFNNC) was proposed to track the MPP. A radial basis function neural network (RBFNN) was developed to provide the reference information to the RFNNC. With a derived learning algorithm, the parameters of the RFNNC were updated adaptively. The mean square error of the estimated tracking error is 0.4×10^-2 which guarantees a good predicting performance of the RBFNN. The RFNNC only needs about 4 ms to reach steady state with small fluctuation. Compared with the fuzzy logic control algorithm, simulation results show that the proposed control algorithm yields much better tracking performance.