arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2501.17296cs.LGcs.AI
收藏

COMPOL:用于可扩展多物理场模拟的统一神经算子框架

COMPOL: A Unified Neural Operator Framework for Scalable Multi-Physics Simulations

School of Information, University of Michigan–Ann Arbor(信息学院,密歇根大学安娜堡分校);Hildebrand Department of Petroleum and Geosystems Engineering, The University of Texas at Austin(石油与地质系统工程系,德克萨斯大学奥斯汀分校);Department of Computer Science, Florida State University(计算机科学系,佛罗里达州立大学)

Junqi Qu, Tao Wang, Yushun Dong, Hewei Tang, Shibo Li

AI总结:

研究针对多物理场模拟中神经算子难以捕捉复杂相关性的问题,提出COMPOL框架,通过循环和注意力聚合机制扩展传统算子架构,实验表明该框架与现有方法相比能实现更高预测精度。

摘要:

多物理场模拟在跨科学和工程领域的复杂相互作用建模中起着至关重要的作用。尽管神经算子尤其是傅里叶神经算子(FNO)显著提高了计算效率,但它们往往无法有效捕捉耦合物理过程中固有的复杂相关性。为解决这一局限性,我们引入了COMPOL,这是一种新颖的耦合多物理场算子学习框架。COMPOL通过纳入复杂的循环和基于注意力的聚合机制扩展了传统算子架构,有效地在潜在特征空间中对相互作用物理过程之间的相互依赖性进行建模。我们的方法与架构无关,并能无缝集成到涉及潜在空间变换的各种神经算子框架中。在包括生物反应扩散系统、模式形成化学反应、多相地质流动和热流体力学过程等各种基准上进行的广泛实验表明,COMPOL与现有方法相比始终能实现更高的预测精度。

英文摘要:

Multiphysics simulations play an essential role in accurately modeling complex interactions across diverse scientific and engineering domains Although neural operators especially the Fourier Neural Operator FNO have significantly improved computational efficiency they often fail to effectively capture intricate correlations inherent in coupled physical processes To address this limitation we introduce COMPOL a novel coupled multiphysics operator learning framework COMPOL extends conventional operator architectures by incorporating sophisticated recurrent and attentionbased aggregation mechanisms effectively modeling interdependencies among interacting physical processes within latent feature spaces Our approach is architectureagnostic and seamlessly integrates into various neural operator frameworks that involve latent space transformations Extensive experiments on diverse benchmarksincluding biological reactiondiffusion systems patternforming chemical reactions multiphase geological flows and thermohydromechanical processes demonstrate that COMPOL consistently achieves superior predictive accuracy compared to stateoftheart methods.

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。

↑