报告题目 Title：建模与仿真的AI革命：赋能计算机感知真实世界 AI Revolution in Modeling and Simulation:Teaching computers to see the real world

报告人 Speaker：蓝楚科 (Timon Rabczuk)

报告人所在单位 Affiliation：复旦大学

时间 Time：2026-04-14 13:30-15:00

地点 Venue：逸夫科技楼一楼报告厅

报告摘要 Abstract：Engineering simulation is at a pivotal juncture. For decades, we have relied on a well-established pipeline-from CAD geometry to mesh generation to physics solvers-to predict the behavior of everything from aircraft wings to the human heart. However, this process is notoriously slow, bottlenecked by manual pre – and post-processing and limited by our ability to calibrate complex models.

This talk will explore how Artificial Intelligence is fundamentally rewriting the rules of simulation. We will journey hrough the modern modeling pipeline, identifying key bottlenecks where AI acts as a powerful accelerant: from automatically repairing CAD geometry and generating meshes to creating ultra-fast surrogate models like Neural Operators and Physics-Informed Neural Networks (PINNs). We will see how Al is not just speeding up existing workflows but enabling entirely new capabilities, such as discovering hidden physical laws from data alone.

Finally, we will look beyond software to the hardware revolution powering this change. We’ll examine how the GPU era enabled the current Al boom and look into the future of computing-from neuromorphic chips to optical solvers-that promise to make real-time, high-fidelity simulation a reality. This transition yields a new paradigm for engineers, merging deep physical intuition with data-driven intelligence to solve tomorrow's most complex challenges. The future of simulation is not just faster; it's intelligent.

个人简介 Bio：蓝楚科（Timon Rabczuk）是计算力学领域国际顶尖学者，复旦大学计算力学与人工智能交叉研究院创始院长。入选欧洲科学院士、欧洲科学与艺术院院士、美国国家人工智能科学院院士。在多尺度与多场耦合力学、生物力学、人工智能驱动力学计算、流体-固体相互作用方面做出了原创性贡献。他同时在“工程”和“计算科学”两个领域被列入ISI全球高被引学者，获德国洪堡基金会费尔多吕楠奖，并担任欧盟ERC、欧盟地平线2020与德国多个大型研究项目首席科学家。蓝楚科院士出版学术专著3部，发表了700多篇国际期刊论文，论文在Web of Science核心引用超过50000次，H指数123。

海报 Poster: 蓝楚科 (Timon Rabczuk) 相辉讲堂.jpg