Improving the Efficiency of the Photovoltaic-Electrolyzer with a new model for Artificial Neural Tracker based on Short Circuit Current Technique

Abstract

This research deals with the improvement of the performance of an alkaline water analyzer system, indirectly coupled with a photovoltaic system. This done by using a neuron tracker based on a neuronal current estimator, and a variable step algorithm for estimating the duty cycle used to adjust the the DC-DC Converter in a Matlab/Simulink environment, to track the maximum power point of the PV system. As a result, providing the largest possible energy to the alkaline the alkaline Electrolyzer, to raise and improve the hydrogen productivity under the available solar irradiance conditions. In this context, the paper presents a new model for a neuronal current estimator, to rapidly and directly estimate the short circuit current, without the need for periodic separation of the PV system from the load to measure the short circuit current, thus avoiding the resulting power loss. So, the optimum operating current for a PV system can be easily determined for tracker MPP  point. The research also proposes algorithm, operating with variable step size, to improve the dynamic performance of the tracker, to achieve the adjustment of the duty cycle of the voltage changer, while tracking the MPP point, and to secure the maximum power transfer to the alkaline Electrolyzer from the PV system. The simulation results of the Photovoltaic-Electrolysis system performed in the Matlab/Simulink environment, showed that the proposed neural Tracker achieves an optimal performance of the operation of the indirectly coupled photoelectric system compared to the direct coupling condition without a tracker and without a voltage converter.