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 2502.04562cs.LGcs.NAmath.NAphysics.flu-dyn
收藏

用于学习边界条件和模型选择的神经算子专家混合

Mixture of neural operator experts for learning boundary conditions and model selection

University at Buffalo, The State University of New York(纽约州立大学布法罗分校);Sandia National Laboratories(桑迪亚国家实验室)

Dwyer Deighan, Jonas A. Actor, Ravi G. Patel

AI总结:

本文提出一种结合体积惩罚与专家混合的神经算子边界条件施加方法,在圆盘和槽道流等非周期域上实现算子学习、模型选择与贝叶斯外推预测。

摘要:

虽然基于傅里叶的神经算子最适合学习周期域上函数之间的映射,但已有若干工作引入了处理非平凡边界条件的技术。然而,所有先前提出的方法都存在限制其适用性的约束。在本工作中,我们引入一种施加边界条件的替代方法,其灵感来自数值方法中的体积惩罚和机器学习中的专家混合(Mixture of Experts, MoE)。通过引入相互竞争的专家,该方法还额外支持模型选择。为演示该方法,我们将空间条件化的 MoE 与基于傅里叶的物理模态算子回归(MOR-Physics)神经算子相结合,并在圆盘和四分之一圆盘上恢复了一个非线性算子。接下来,我们从槽道流的直接数值模拟(DNS)中提取大涡模拟(LES)模型,并展示我们的方法所提供的域分解。最后,我们用贝叶斯变分推断训练该 LES 模型,并获得远超出 DNS 模拟时间范围之后的流动后验预测样本。

英文摘要:

While Fourier-based neural operators are best suited to learning mappings between functions on periodic domains, several works have introduced techniques for incorporating non trivial boundary conditions. However, all previously introduced methods have restrictions that limit their applicability. In this work, we introduce an alternative approach to imposing boundary conditions inspired by volume penalization from numerical methods and Mixture of Experts (MoE) from machine learning. By introducing competing experts, the approach additionally allows for model selection. To demonstrate the method, we combine a spatially conditioned MoE with the Fourier based, Modal Operator Regression for Physics (MOR-Physics) neural operator and recover a nonlinear operator on a disk and quarter disk. Next, we extract a large eddy simulation (LES) model from direct numerical simulation of channel flow and show the domain decomposition provided by our approach. Finally, we train our LES model with Bayesian variational inference and obtain posterior predictive samples of flow far past the DNS simulation time horizon.

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。

↑