An ANN-INC MPPT Strategy for Photovoltaic System

Abstract

Maximum power point tracking (MPPT) technology is widely used to achieve high-efficiency output of photovoltaic system under the changing irradiation and temperature conditions. To improve the efficiency of the photovoltaic system, a MPPT control strategy based on neural network and incremental conductance (ANN-INC) is proposed in this paper. In ANN-INC, the output of the trained neural network is transferred to the INC part as the initial duty cycle, which makes the initial duty cycle have a small gap with the duty cycle when the photovoltaic system works at MPP, then the INC part can select a small step size to make the output of the photovoltaic system closer to expected output. The strategy has simple structure, fast dynamic response speed, small steady-state power oscillations and high efficiency. The strategy also performs well when the irradiation changes rapidly. The superiority of the strategy is verified in Matlab / Simulink.