Photovoltaic Maximum Power Point Tracking Control and Neural Network Modeling Based on Improved Perturbation Observation Method.

Solar energy is one of the most widely used clean energy in human society. In order to improve the photoelectric conversion efficiency of solar energy, a variety of maximum power tracking methods for photovoltaic (PV) cells are proposed. Perturbation observation method (P&O) is one of the most commonly used methods for maximum power tracking. However, the conventional P&O cannot balance the tracking speed and tracking accuracy. Based on this, an improved P&O is proposed in this paper. At the same time, in order to facilitate the power tracking and control of PV cells system, a state-space oriented model of PV cells MPPT system is established in this paper, and recursive neural network (RNN) is used to complete the identification of model parameters. Simulation results show that the proposed P&O and the established state space model are effective. The results show that the improved P&O proposed in this paper can achieve rapid and accurate maximum power tracking of PV cells. At the same time, the state space model of MPPT system based on RNN has higher approximation accuracy to the original model.