题目： Sparse Operator Learning and Multi-continuum Modeling for Multiscale Problems

主讲人：王亚婷 副教授

时间：2025年5月19日（周一）10：30

地点：理学院楼308

主办单位：理学院

主讲人简介：

王亚婷，西安交通大学副教授。2018年获美国Texas A&M University博士学位，2018-2021年在普渡大学从事博士后研究工作，2021-2022年为香港大学数学系tenure-track助理教授，2022年7月起任职于西安交通大学数学与统计学院，主要从事多尺度模型降阶和相关的深度学习方法研究。已在SISC、JCP等计算数学主流期刊上发表学术论文20余篇，主持国自然青年项目一项，获陕西省科协青年人才托举计划支持。

摘要：

     The inherent multiscale and uncertainty properties in physical systems present substantial challenges for numerical simulations, making it important to develop robust reduced-order models with efficient computational implementations. To address highly heterogeneous multiscale phenomena exhibiting high contrast ratios and non-local features, we first present some localized model reduction methods. Building upon these foundational techniques, we then develop a sparse operator learning framework that systematically integrates data with physical knowledge. On the other hand, to bridge the persistent gap between numerical predictions and empirical observations, we further assimilate data into the poor reduced order model, obtaining an improved multi-continuum model that better align with experimental data and physical mechanism using deep learning approach.