“拓界”读书会｜第一季“Deep AI+” 第七期讲座预告

第七期讲座精彩导览

“拓界”读书会第一季Deep AI+ 第七期讲座

人工智能在有限元计算中的应用

Application of artificial intelligence in finite element calculation

2022年12月3日（星期六）

下午15:00—16:30

腾讯会议：417-501-554

会议密码：4169

15:00 pm—16:30 pm, 3rd Dec. 2022 (Sat.)

Tencent Meeting ID: 417 501 554

Password: 4169

主讲人：孟诗乔

同济大学土木工程学院结构防灾减灾工程系 21 级直博生，曾获得同济大学追求卓越学生提名奖、同济大学本科生学术之星、上海市优秀毕业生等荣誉称号。本科期间参加学科竞赛获奖 20余项，包括国际级两项、国家级五项、省部级九项涵盖数学、力学、计算机等多个学科领域。

主要研究方向为人工智能和土木工程的学科交叉方向。

已在 Computer-Aided Civil and Infrastructure Engineering 等高水平 SCI一区TOP期刊上以第一作者发表论文三篇，发表 EI 论文两篇，申请发明专利三项。

He is a direct doctoral student of the Department of Structural Disaster Prevention and Mitigation Engineering, School of Civil Engineering, Tongji University. He has been awarded the nomination award for Pursuit of Excellence, Academic Star of Tongji University, and Excellent Graduate of Shanghai. During his undergraduate years, he won more than 20 academic competitions, including two international, five nationals, and nine provincial and ministerial level competitions in various fields such as mathematics, mechanics, and computer science. 

His main research direction is the interdisciplinary direction of artificial intelligence and civil engineering. 

He has published three papers as the first author in Computer-Aided Civil and Infrastructure Engineering, other top SCI journals, and two EI papers and applied for three invention patents.

主持人：张俊豪

Host：junhao Zhang

报告内容介绍

有限元法是一种为求解偏微分方程边值问题近似解的数值技术，是一种常用的工程分析手段，已成为各种物理过程的工程设计分析和科学建模的计算主力。尽管有限元计算可以在满足工程要求的精度下实现多数问题的求解，其较大的算力需求和较长的计算时间仍是一个巨大的问题。近年来，随着机器学习的高速发展以及多种基于物理信息的神经网络的提出，深度学习在科学计算和知识发现中的应用日渐火热。本次报告首先介绍有限元计算的理论基础和求解方法；之后介绍几种经典的应用深度学习求解微分方程的框架；最后介绍近期的将深度学习应用于有限元计算中从而获得计算速度提升的相关工作。

The finite element method is a numerical technique for solving approximate solutions of partial differential equation boundary value problems. Although finite element calculation can solve most problems with the accuracy required by engineering, its large computing power requirement and long computing time are still huge problems. In recent years, with the rapid development of machine learning and the proposal of various neural networks based on physical information, deep learning in scientific computing and knowledge discovery has become increasingly popular. This report first introduces the theoretical basis and solution methods of finite element calculations, then introduces several classic frameworks for applying deep learning to solve differential equations. Finally, this report introduces the related research on applying deep learning to finite element calculations to improve calculation speed.

大纲

1   有限元计算方法

Finite element calculation method

2   基于深度学习的微分方程求解框架
Framework for solving differential equations based on deep learning

3   深度学习方法在有限元计算中的应用

Application of deep learning method in finite element calculation

关键词

有限元计算、物理引导的机器学习、数据驱动的发现、多尺度模拟

Finite element calculation, Physics-guided machine learning, Data-driven discovery;, Multiscale simulation

参考文献

[1] Saha S, Gan Z, Cheng L, et al. Hierarchical deep learning neural network (HiDeNN): An artificial intelligence (AI) framework for computational science and engineering[J]. Computer Methods in Applied Mechanics and Engineering, 2021, 373: 113452.

[2] Haghighat E, Raissi M, Moure A, et al. A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics[J]. Computer Methods in Applied Mechanics and Engineering, 2021, 379: 113741.

[3] Raissi M, Perdikaris P, Karniadakis G E. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J]. Journal of Computational physics, 2019, 378: 686-707.

[4] Zhang L, Cheng L, Li H, et al. Hierarchical deep-learning neural networks: finite elements and beyond[J]. Computational Mechanics, 2021, 67(1): 207-230.

[5] 王勖成. 有限单元法[M]. 清华大学出版社. 2003.

       【关于我们】

     “拓界”读书会是在校团委指导下，由同济大学学生科学技术协会组织创建的跨学科共学社群。我们的目的是，构建拓展视野、宏观思维、交流知识、提升生活的社群；我们的形式是，围绕若干研究热点/前沿领域/学术论文、以线上线下相结合的方式开展读书会；我们涉及的学科领域包括但不限于：人工智能、医学、社科等。

     第一季“拓界”读书会的主题为“Deep AI+”，希望能通过本季读书会，一起探讨和交流“How to make AI+ deeper”。能够让做AI理论研究的老师和同学们理解传统行业场景和深入了解基础理论研究，并且能够让大家理性认知AI工具：它们既不是万能钥匙，也不是毫无理论支持的拟合工具。

                                        【加入我们】

     如您想获取更多关于读书会的消息，请联系管理员入群（管理员微信号：zzzzzzet1；备注信息请填：加入读书会总群）。或者关注“同济学生科协”公众号下的读书会栏目。

     如您想加入我们，成为下一季主题发起人或主讲人，可以联系以下负责人：

潘   琪 土木工程学院博士研究生 

手机号/微信号：13135530820

高翔宇 数学科学学院博士研究生

手机号：[phone] / 微信号：CH3COCl

编辑｜马浩洋 潘琪

校审｜潘琪 霍钧资 张绣宇 孙羽捷

海报设计｜屈泊彤 蒋天逸 霍钧资