Скачать или смотреть Huan Lei - Energy-stable machine-learning model of non-Newtonian hydrodynamics w/ molecular fidelity

Recorded 08 October 2025. Huan Lei of Michigan State University presents "An energy-stable machine-learning model of non-Newtonian hydrodynamics with molecular fidelity" at IPAM's Bridging Scales from Atomistic to Continuum in Electrochemical Systems Workshop.
Abstract: One essential challenge in the computational modeling of multi-scale systems is the availability of reliable and interpretable closures that faithfully encode the micro-scale interactions. For systems without clear scale separation, there generally exists no such a simple set of macro-scale field variables that allow us to project and predict the dynamics in a self-determined way. We introduce a machine-learning (ML) based approach that enables us to reduce high-dimensional multi-scale systems to truly reliable macro-scale models with low-dimensional variational structures that preserve canonical degeneracies and symmetry constraints. The non-Newtonian hydrodynamics of polymeric fluids is used as an example to illustrate the essential idea. Unlike our conventional wisdom about ML modeling that focuses on learning the partial differential equation (PDE) form, the present approach directly learns the energy variational structure from the micro-model through an end-to-end process via the joint learning of a set of micro-macro encoder functions. The final model retains a multi-scale nature with clear physical interpretation. Various pre-existing energy stable numerical schemes can be naturally used, which ensures the computational efficiency and numerical robustness for real applications.
Learn more online at: https://www.ipam.ucla.edu/programs/wo...