Simple computational strategies for more effective physics-informed neural networks modeling of turbulent natural convection

Didier Lucor (1, 2) , Atul Agrawal , Anne Sergent

1 LISN - Laboratoire Interdisciplinaire des Sciences du Numérique
2 DATAFLOT - DAtascience, trAnsition, Fluid instability, contrOl, Turbulence

Domains: Fluids mechanics [physics.class-ph]; Computational Engineering, Finance, and Science [cs.CE]; Data Analysis, Statistics and Probability [physics.data-an]; Artificial Intelligence [cs.AI]

Didier Lucor, Atul Agrawal, Anne Sergent. Simple computational strategies for more effective physics-informed neural networks modeling of turbulent natural convection. Journal of Computational Physics, 2022, 456, pp.111022. ⟨10.1016/j.jcp.2022.111022⟩. ⟨hal-03847907⟩