Research on the Prediction of Extreme Refueling Conditions for the Fluid–Solid Coupling Refueling Process of Hydrogen-Storage Cylinders in Fuel-Cell Vehicles

In the industrialization process of fuel-cell vehicles, the safety and efficiency of gaseous hydrogen refueling for on-board hydrogen-storage cylinders (HSCs) have always been critical bottlenecks awaiting breakthroughs. However, the vague definition of boundaries for refueling conditions in existing theories has become a prominent challenge restricting the formulation of specifications and industrial applications. To solve the aforementioned problem, a numerical model for gaseous hydrogen refueling considering the fluid–solid coupling of HSCs was established, and a prediction model for extreme refueling conditions was constructed by integrating machine learning. The results show that the coefficient of determination ( R 2 ) of this prediction model reaches 0.99, indicating excellent prediction accuracy. Further analysis reveals that the extreme inlet temperature has a linear positive correlation with the initial pressure, the extreme refueling rate exhibits a quadratic negative correlation with the inlet temperature, and the extreme initial pressure shows a cubic positive correlation with the refueling rate. The optimized extreme refueling conditions based on the aforementioned laws can increase the maximum refueling efficiency by 8.33%, reduce the precooling energy consumption by 12%, and expand the applicable pressure range by 2.97%, under the premise that the temperature of the HSC does not exceed the safety limit. The research findings provide new ideas and references for the scientific formulation of safety specifications for gaseous hydrogen refueling.