Qiang Liu

Qiang Liu | 刘强

I am a PhD student supervised by Prof. Nils Thuerey at the Technical University of Munich . My research interest focuses on physics-based deep learning methods and advanced technologies for fluid simulations.

Currently, I am working on leveraging the Denoising Diffusion Probabilistic Model (DDPM) for physical applications. DDPM has emerged as a state-of-the-art generative model, demonstrating superior performance in synthesizing impressive results across various domains. My study emphasizes utilizing DDPM’s capability to reconstruct diverse probability distributions accurately for studying physical problems involving uncertainty. Meanwhile, by integrating existing physical knowledge ⚛️ into the training of the diffusion model, I aim to elevate DDPM into a cutting-edge generative model for complex physical systems.

I use to work with Prof Jian Wu on Electrohydrodynamics at Harbin Institute of Technology .

Featured Publications

AIAAJ
Uncertainty-aware Surrogate Models for Airfoil Flow Simulations with Denoising Diffusion Probabilistic Models

Qiang Liu , and Nils Thuerey

AIAA Journal, 2024

Abs Bib HTML Blog Code

Leveraging neural networks as surrogate models for turbulence simulation is a topic of growing interest. At the same time, embodying the inherent uncertainty of simulations in the predictions of surrogate models remains very challenging. The present study makes a first attempt to use denoising diffusion probabilistic models (DDPMs) to train an uncertainty-aware surrogate model for turbulence simulations. Due to its prevalence, the simulation of flows around airfoils with various shapes, Reynolds numbers, and angles of attack is chosen as the learning objective. Our results show that DDPMs can successfully capture the whole distribution of solutions and, as a consequence, accurately estimate the uncertainty of the simulations. The performance of DDPMs is also compared with varying baselines in the form of Bayesian neura networks and heteroscedastic models. Experiments demonstrate that DDPMs outperformthe other methods regarding a variety of accuracy metrics. Besides, it offers the advantageof providing access to the complete distributions of uncertainties rather than providing a set of parameters. As such, it can yield realistic and detailed samples from the distribution of solutions.
@article{airfoilDDPM2023, journal = {AIAA Journal}, title = {Uncertainty-aware Surrogate Models for Airfoil Flow Simulations with Denoising Diffusion Probabilistic Models}, author = {Liu, Qiang and Thuerey, Nils}, year = {2024}, eprint = {2312.05320}, archiveprefix = {arXiv}, primaryclass = {physics.flu-dyn}, arixv = {https://arxiv.org/abs/2312.05320}, url = {https://arc.aiaa.org/doi/10.2514/1.J063440}, doi = {10.2514/1.J063440}, volume = {62}, issue = {8}, pages = {2192-2933}, }
Arxiv
ConFIG: Towards Conflict-free Training of Physics Informed Neural Networks

Qiang Liu , Mengyu Chu , and Nils Thuerey

Arxiv preprint, 2024

Abs Bib HTML Blog Code

The loss functions of many learning problems contain multiple additive terms that can disagree and yield conflicting update directions. For Physics-Informed Neural Networks (PINNs), loss terms on initial/boundary conditions and physics equations are particularly interesting as they are well-established as highly difficult tasks. To improve learning the challenging multi-objective task posed by PINNs, we propose the ConFIG method, which provides conflict-free updates by ensuring a positive dot product between the final update and each loss-specific gradient. It also maintains consistent optimization rates for all loss terms and dynamically adjusts gradient magnitudes based on conflict levels. We additionally leverage momentum to accelerate optimizations by alternating the back-propagation of different loss terms. The proposed method is evaluated across a range of challenging PINN scenarios, consistently showing superior performance and runtime compared to baseline methods. We also test the proposed method in a classic multi-task benchmark, where the ConFIG method likewise exhibits a highly promising performance.
@article{ConFIG2024, journal = {Arxiv preprint}, title = {ConFIG: Towards Conflict-free Training of Physics Informed Neural Networks}, author = {Liu, Qiang and Chu, Mengyu and Thuerey, Nils}, year = {2024}, eprint = {2408.11104}, archiveprefix = {arXiv}, primaryclass = {cs.LG}, arixv = {https://arxiv.org/abs/2408.11104}, url = {https://arxiv.org/abs/2408.11104}, doi = {https://doi.org/10.48550/arXiv.2408.11104}, }

Featured Projects

ConvDO: Convolutional Differential Operators for Physics-based Deep Learning Study