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 2603.22309cs.LGcs.AI
收藏

UniFluids: 基于条件流匹配的统一神经算子学习

UniFluids: Unified Neural Operator Learning with Conditional Flow-matching

Academy of Mathematics and Systems Science, Chinese Academy of Sciences(中国科学院数学与系统科学研究院);Microsoft Research Asia(微软亚洲研究院);Renmin University of China(中国人民大学);Institute of Zoology, Chinese Academy of Sciences(中国科学院动物研究所)

Haosen Li, Qi Meng, Jiahao Li, Rui Zhang, Ruihua Song, Liang Ma, Zhi-Ming Ma

AI总结:

本文提出UniFluids框架,利用扩散Transformer的可扩展性,统一学习不同维度和物理变量的PDE解算子,通过四维时空表示实现联合训练与条件编码,提升预测精度与泛化能力。

摘要:

偏微分方程(PDE)模拟在科学研究中具有重要价值。目前,利用深度神经网络学习PDE解算子具有巨大潜力。本文提出UniFluids,一种基于条件流匹配的框架,利用扩散Transformer的可扩展性,统一学习不同维度和物理变量的PDE解算子。与自回归PDE基础模型不同,UniFluids采用流匹配实现并行序列生成,成为首个统一算子学习的方法。具体而言,引入统一的四维时空表示用于异构PDE数据集,实现联合训练和条件编码。此外,发现PDE数据集的有效维度远低于其块维度,因此在流匹配算子学习中采用$x$-预测,验证其显著提升预测精度。在多个涵盖1D、2D和3D空间维度的PDE数据集上进行大规模评估,实验结果表明UniFluids在预测精度、可扩展性和跨场景泛化能力方面表现优异。代码将在后续发布。

英文摘要:

Partial differential equation (PDE) simulation holds extensive significance in scientific research. Currently, the integration of deep neural networks to learn solution operators of PDEs has introduced great potential. In this paper, we present UniFluids, a conditional flow-matching framework that harnesses the scalability of diffusion Transformer to unify learning of solution operators across diverse PDEs with varying dimensionality and physical variables. Unlike the autoregressive PDE foundation models, UniFluids adopts flow-matching to achieve parallel sequence generation, making it the first such approach for unified operator learning. Specifically, the introduction of a unified four-dimensional spatiotemporal representation for the heterogeneous PDE datasets enables joint training and conditional encoding. Furthermore, we find the effective dimension of the PDE dataset is much lower than its patch dimension. We thus employ $x$-prediction in the flow-matching operator learning, which is verified to significantly improve prediction accuracy. We conduct a large-scale evaluation of UniFluids on several PDE datasets covering spatial dimensions 1D, 2D and 3D. Experimental results show that UniFluids achieves strong prediction accuracy and demonstrates good scalability and cross-scenario generalization capability. The code will be released later.

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。

↑