Comparative study of photovoltaic system for hydrogen electrolyzer system

Abstract

Energy storage remains a strategic challenge in energy control. The world is turning to decentralized generators, including fuel cell systems, which use hydrogen as fuel. This gives a special interest to the system of storage of the hydrogen produced mainly by the solar energy. The electrolyzer becomes the most prominent device, but its disadvantage is that it requires a lot of electrical energy. This present paper deals with controlling the power emanating from the photovoltaic system and consumed by the electrolyzer, through the use of neural networks to optimize the power generated by the photovoltaic system. In this work, we will compare between two types of converters Cuk and SEPIC because they are the most widely used and are two of the developed family of the converter. This paper presents under MATLAB/Simulink the use of Cuk converter with maximum power point tracking (MPPT) technology, to increase its efficiency by an algorithm Perturb and Observe (P&O) and incremental conductance method, then we will apply artificial neural network (ANN) to avoid the disadvantages of MPPT Classical. The MPPT developed with the artificial neural networks presents a better behavior than the classic system Perturb & Observe.