| 研究開始時の研究の概要 |
We will develop Machine-Learning assisted simulation methods for (A) entangled polymer melt flows and (B) Stokes flows (i.e., flows at small scales). For (A), we will infer the relationship between microscopic and macroscopic degrees of freedom, as well as optimize the flow. For (B) we will develop a probabilistic framework to infer the flow solution in complex environments, with moving boundaries, from partial/noisy data. These methods will be used to analyze (A) industrial polymer processing flows, as well as (B) Stokes flows typical of biological or colloidal systems.
|
