arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2212.00716physics.flu-dyncs.LGcs.NAmath.NA
收藏

使用梯度提高流体动力学深度学习模型的泛化性能

Using Gradient to Boost the Generalization Performance of Deep Learning Models for Fluid Dynamics

MINES ParisTech(巴黎高科矿业学院);PSL University(巴黎文理研究大学);Extrality

Eduardo Vital Brasil

摘要:

当今,计算流体动力学(CFD)是工业设计的基本工具。然而,进行此类模拟的计算成本很昂贵,并且对于需要进行许多模拟的现实用例(例如形状优化任务)来说可能是有害的。最近,深度学习(DL)在广泛的应用中取得了重大飞跃,成为物理系统的良好候选者,为CFD开辟了前景。为了规避CFD的计算瓶颈,DL模型已被用于在欧几里德数据上学习,最近,在非欧几里德数据上学习,如非结构化网格和流形,允许更快,更有效的(内存,硬件)代理模型。然而,DL提出了从训练数据分布(设计空间)外推(推广)的内在限制。在这项研究中,我们提出了一项新的工作,以提高深度学习的泛化能力。为此,我们将物理梯度(输出w.r.t.输入)到DL模型。我们的策略已经显示出良好的结果,更好地推广DL网络和我们的方法/理论研究与实证验证,包括消融研究证实。

英文摘要:

Nowadays, Computational Fluid Dynamics (CFD) is a fundamental tool for industrial design. However, the computational cost of doing such simulations is expensive and can be detrimental for real-world use cases where many simulations are necessary, such as the task of shape optimization. Recently, Deep Learning (DL) has achieved a significant leap in a wide spectrum of applications and became a good candidate for physical systems, opening perspectives to CFD. To circumvent the computational bottleneck of CFD, DL models have been used to learn on Euclidean data, and more recently, on non-Euclidean data such as unstuctured grids and manifolds, allowing much faster and more efficient (memory, hardware) surrogate models. Nevertheless, DL presents the intrinsic limitation of extrapolating (generalizing) out of training data distribution (design space). In this study, we present a novel work to increase the generalization capabilities of Deep Learning. To do so, we incorporate the physical gradients (derivatives of the outputs w.r.t. the inputs) to the DL models. Our strategy has shown good results towards a better generalization of DL networks and our methodological/ theoretical study is corroborated with empirical validation, including an ablation study.

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。

↑