Research progress of porous material permeability prediction based on deep learning
As an important parameter to describe the flow characteristics of porous materials,permeability is wildly used in various fields such as mechanics and energy.Empirical formulas is the traditional methods for predicting per-meability in porous materials,and have some limitations in terms of generality and prediction accuracy.In recent years,there has been growing interest among researchers in utilizing deep learning to construct permeability predic-tion models,which shows promising prospects in addressing the shortcomings of empirical formulas.Focusing on the modeling approach for permeability prediction in porous materials based on deep learning,the application and de-velopment trends of deep learning techniques in porous material reconstruction were discussed.An overview of the fundamental principles and research progress was addressed for fast prediction modeling methods of permeability based on the relationships between permeability and structural parameters,permeability and images,as well as per-meability and flow field.Finally,future research directions in this field and the potential for enhancing the perform-ance of porous material manufacturing systems were discussed.
deep learningporous materialmodel reconstructionpermeabilityprediction model
NETL
NSTL
万方数据
