Artificial Intelligence Based Forecasting Models for Solar PV/Hydrogen Fuel Cell Power System: Role of Short-Term AI-Based Forecasting for the Energy Transition

This study aims to assess the viability of using solar photovoltaic (PV) water electrolyzer to generate green hydrogen and sustainable energy, with the specific goal of meeting the electrical demands of buildings in arid locations. Models for short-term ANN power output predictions from solar PV and fuel cells, as well as green hydrogen production, are developed. By utilizing a power system fueled exclusively by renewable resources and zero-carbon fuels, our objective is to build a structure that operates without consuming fossil fuel energy. Modeling, simulation, and optimization analysis were used in this study to evaluate the system's technical, economic, and environmental performance. The artificial neural network (ANN) prediction models utilize simulated input and output data to forecast the production of hydrogen through solar-powered electrolyzer. The building's daily, monthly, and annual electricity demands were all fulfilled by the self-sufficient hybrid solar PV/fuel cell power system. The hybrid power system being proposed generates 79.7% and 20.3% of its electricity from solar photovoltaic (PV) and fuel cell sources, respectively. The electrolyzer requires an energy input of 898,415 kilowatt-hours, produces 19,360 kilograms of hydrogen, and consumes 15,817 kilograms of hydrogen for the fuel cell per year. The mean and maximum output of the electrolyzer for green hydrogen production are 2.21 and 10.8 kg/hr., respectively, with a specific energy of 46.4 kWh/kg. The training, validation, and test R values for solar PV power output and green hydrogen production, obtained from the ANN-based forecasting models, range from 0.967 to 0.974 and from 0.981 to 0.982, respectively. The correlation coefficient R and mean square errors MSE of the ANN regression model indicate that the models provide accurate estimations of the power output of solar PV systems and the production of green hydrogen by electrolyzer.