arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.03627cs.LG
收藏

通过物理引导的少步扩散实现超快PDE求解

Ultra Fast PDE Solving via Physics Guided Few-step Diffusion

Purdue University(普渡大学);hi-Lab, Xiaohongshu Inc(小红书实验室,小红书公司)

Cindy Xiangrui Kong, Yueqi Wang, Haoyang Zheng, Weijian Luo, Guang Lin

AI总结:

Phys-Instruct通过物理引导的蒸馏框架,实现PDE求解的高效与高物理一致性。

摘要:

基于扩散模型的PDE求解方法在准确性和泛化性方面表现出色,但仍然面临显著的局限性,如高采样成本和物理一致性不足,这些限制源于其多步迭代采样机制和缺乏显式的物理约束。为了解决这些问题,我们提出了Phys-Instruct,一种新颖的物理引导的蒸馏框架,它不仅(1)通过匹配生成器和先验扩散分布将预训练的扩散PDE求解器压缩成一个少步生成器,以实现快速采样,而且(2)通过通过PDE蒸馏引导显式注入PDE知识来增强物理一致性。Phys-Instruct建立在坚实的理论基础上,导致一个具有可计算梯度的实用物理约束训练目标。在五个PDE基准测试中,Phys-Instruct在推理速度上实现了数量级的提升,同时将PDE误差降低了超过8倍,相较于最先进的扩散基线。此外,所得到的无条件学生模型作为紧凑的先验,能够为各种下游条件任务提供高效且物理一致的推理。我们的结果表明,Phys-Instruct是一种由深度生成模型驱动的新型、有效且高效的超快PDE求解框架。

英文摘要:

Diffusion-based models have demonstrated impressive accuracy and generalization in solving partial differential equations (PDEs). However, their practical utility still suffers from high sampling costs and insufficient physical consistency, mainly due to many-step iterative sampling and insufficient physics guidance. To address these issues, we propose \emph{Phys-Instruct}, a physics-guided distillation framework that compresses a pre-trained diffusion PDE solver into a few-step generator while explicitly enforcing PDE consistency through distillation guidance. We reformulate the training objective as a constrained problem with PDE residual, and derive a score-function identity that leads to a practical physics-guided generator update. Across five PDE benchmarks, Phys-Instruct achieves orders-of-magnitude faster inference while reducing the PDE Error by factors of $2.4$ to $15.5$, compared with the corresponding 200-step EDM teachers. Moreover, the resulting unconditional student model functions as a compact prior, enabling efficient and physically consistent inference for various downstream conditional tasks. Our results indicate that Phys-Instruct is an effective and efficient framework for ultra-fast PDE solving powered by generative models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑