Lastest News

Invited seminar talk at Cornell on Neural Differentiable Physics


Title: Neural Differentiable Physics: Unifying Numerical PDEs and Deep Learning for Data-Augmented Computational Physics


Predictive modeling and simulation are essential for understanding, predicting, and controlling complex physical processes across many engineering disciplines. However, traditional numerical models, which are based on first principles, face significant challenges, especially for complex systems involving multiple interacting physics across a wide range of spatial and temporal scales. (1) A primary obstacle stems from our often-incomplete understanding of the underlying physics, which results in inadequate mathematical models that fail to accurately capture system behavior. (2) Additionally, the high computational demands of traditional solvers represent another substantial barrier, especially when real-time control or many repeated model queries are required, as in design optimization, inference, and uncertainty quantification. Fortunately, the continual evolution of sensing technology and the exponential increase in data availability have opened new avenues for the development of data-driven computational modeling frameworks. Bolstered by advanced machine learning and GPU computing techniques, these models hold the promise of greatly enhancing our predictive capabilities, effectively tackling the challenges posed by traditional numerical models. While data science and machine learning offer novel methods for computational mechanics models, challenges persist, such as the need for extensive data, limited generalizability, and lack of interpretability. Addressing existing challenges for predictive modeling issues requires innovative computational methods that integrate advanced machine learning techniques with physics principles. This talk will introduce some of our efforts along this direction, spotlighting the Neural Differentiable Physics, a novel SciML framework unifying classic numerical PDEs and advanced deep learning models for computational modeling of complex physical systems. Our approach centers on the integration of numerical PDE operators into neural architectures, enabling the fusion of prior knowledge of known physics, multi-resolution data, numerical techniques, and deep neural networks through differentiable programming. The way for integrating physics into the deep learning model represents a novel departure from existing SciML frameworks, such as Physics-Informed Neural Networks (PINNs). By combining the strengths of known physical principles and established numerical techniques with cutting-edge deep learning and AI technology, this innovative framework promises to inaugurate a new era in the understanding and modeling of complex physical systems, with far-reaching implications for science and engineering applications.

Our group will give 6 presentations @ APS DFD 2023

Our group, CoMSAIL, will be delivering 6 presentations at APS DFD 2023. If you are also attending the conference in DC, please stop by and check out our talks (focusing on differentiable physics, GPU computing, and hybrid neural modeling for fluid flow)

J29.00003, Fan et al. Differentiable vectorized JAX solver for turbulent flow and fluid-structure interactions (4:35 PM–6:32 PM, Sunday, November 19, 2023, Session J 29, Room: 152B)

L17.00003 Sun et al. Unifying Predictions of Deterministic and Stochastic Physics in Mesh-reduced Space with Sequential Flow Generative Model (8:00 AM–10:36 AM, Monday, November 20,
Room: 145B)

L17.00005, Pan et al. Neural field based sequence model for generating spatial-temporal turbulence (8:00 AM–10:36 AM, Monday, November 20, 2023, Room: 145B)

X29.00001, Akhare et al. Implicit Neural Solver for Stable Surrogate Simulation of Fluid Dynamics (8:00 AM–10:36 AM, Tuesday, November 21, 2023, Room: 152B)

X29.00006, Liu et al. MuRFiV-Net: A Multi-Resolution Finite-Volume Inspired Neural Network for Predicting Spatiotemporal Dynamics (8:00 AM–10:36 AM, Tuesday, November 21, 2023, Room: 152B)

X29.00011, Zhang et al. A Differentiable Hybrid Neural Solver for Efficient Simulation of Cavitating Flows (8:00 AM–10:36 AM, Tuesday, November 21, 2023, Room: 152B)

Also one collaborative work with Prof. X. Zheng from RIT — R14.00007, Zheng et al. A hybrid physics informed neural network model for patient specific phonation simulation (1:50 PM–3:21 PM, Monday, November 20, 2023, Room: 144AB)

Our lab received research grant from Lucy Institute

We got one-year research grant from Lucy Institute of Data Science and Society.

Jian-Xun Wang, Assistant Professor in the Department of Aerospace and Mechanical Engineering (AME), and Chaoli Wang, Professor in CSE, were awarded this grant to develop a novel Bayesian deep learning framework for automating vessel segmentation from medical images to enable rapid construction of personalized, patient-specific computational models.

Thank Lucy Institute!

Welcome submissions to article collection of “Physics-informed Machine Learning”

We are welcoming submission of original research to an article collection entitled “Physics-informed Machine Learning and Its Real-world Applications” of Scientific Reports (Springer Nature). As Guest Editor for the Collection, I hope you will consider this excellent outlet for your research in this field.

This article collection is a great opportunity to highlight this important area, and we hope you will be able to contribute. Please don’t hesitate to contact me if you have any questions. The deadline is Nov. 30, 2022. For more details, please see