Application of machine learning in modelling gas dispersion coefficients for hydrogen storage in porous media

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Research demonstrates improved gas dispersion coefficient predictions in hydrogen storage, suggesting better cushion gas selection.

Support vector regression achieved high accuracy in predicting gas dispersion coefficients under realistic conditions.
The model recorded R² values above 0.996 for methane, nitrogen, and carbon dioxide, indicating excellent predictive performance.
Laboratory experiments combined with machine learning allowed for capturing the nonlinear behavior of gas mixing in porous media.
These insights enhance hydrogen storage reliability and contribute to sustainable energy systems.

Ali Akbari (2025) studied this question.

View Full Paper

Ask AI

Bookmark