Artificial Neural-Network-Based Maximum Power Point Tracking for Photovoltaic Pumping System Using Backstepping Controller

This paper investigates a comparative analysis of control methods to achieve an optimal photovoltaic (PV) module output voltage and to track the maximum power point (MPP) of the PV system under variable conditions such as temperature, solar irradiance, and load changes, with the help of an offline trained actificial neural (ANN) network. The trained ANN provides the reference voltage corresponding to the MPP for the feed-forward loop which is responsible for regulation of the solar cell array voltage in MPP. The PV system consists of a solar module and a boost DC/DC converter connected to a DC motor which feeds a centrifugal pump for water pumping. Depending on the voltage error signal, the controllers generate a control signal for the pulse-width modulation (PWM) generator which in turn adjusts the duty cycle of the converter. For this purpose first, the proportional-integral (PI) controller is used. Next controllers are based on backstepping approach and the backstepping with integral action. Different simulation tests using Maltab/Simulink environment are given to demonstrate the efficiency of the controllers in presence of the irradiance perturbations.