Skip to main content
QUICK REVIEW

[论文解读] DeepONet prediction of linear instability waves in high-speed boundary layers

Patricio Clark Di Leoni, Lipeng Lu|arXiv (Cornell University)|May 18, 2021
Fluid Dynamics and Turbulent Flows参考文献 1被引用 30
一句话总结

本文训练 DeepONet 来学习在可压边界层中支配线性不稳定波的算子,实现快速下游预测与从壁面数据的反演重构,并具有数据同化应用。

ABSTRACT

Deep operator networks (DeepONets) are trained to predict the linear amplification of instability waves in high-speed boundary layers and to perform data assimilation. In contrast to traditional networks that approximate functions, DeepONets are designed to approximate operators. Using this framework, we train a DeepONet to take as inputs an upstream disturbance and a downstream location of interest, and to provide as output the perturbation field downstream in the boundary layer. DeepONet thus approximates the linearized and parabolized Navier-Stokes operator for this flow. Once trained, the network can perform predictions of the downstream flow for a wide variety of inflow conditions, without the need to calculate the whole trajectory of the perturbations, and at a very small computational cost compared to discretization of the original equations. In addition, we show that DeepONets can solve the inverse problem, where downstream wall measurements are adopted as input and a trained network can predict the upstream disturbances that led to these observations. This capability, along with the forward predictions, allows us to perform a full data assimilation cycle: starting from wall-pressure data, we predict the upstream disturbance using the inverse DeepONet and its evolution using the forward DeepONet.

研究动机与目标

  • 促进在高速边界层中对线性不稳定波放大进行准确、高效的预测,以帮助过渡预测与控制。
  • 开发一个 DeepONet 框架,将支配的动力学视为一个算子,将上游扰动映射到下游扰动场。
  • 证明对不同入口扰动的下游流场正向预测具有低计算成本。
  • 展示逆向能力:从下游壁面测量推断上游扰动,以实现数据同化。

提出的方法

  • 采用 DeepONet 架构,分支网络处理离散化的输入函数,主干网络处理下游的评估点。
  • 建立将上游扰动 f 映射到下游扰动 G(f)(ζ) 的算子 G†,并通过损失 L 比较 G(f)(ζ) 与 G†(f)(ζ) 来训练网络以近似该算子。
  • 训练多个 DeepONets 来模拟描述在发展中的边界层中线性化扰动演化的抛物型稳定性方程(PSE)。
  • 使用正向模型从上游扰动预测下游场(如速度、压力、温度),并用反向模型从下游壁压数据恢复上游扰动。
  • 通过相位和频率变化进行数据增强,并应用加权损失以在衰减的下游模态之间平衡幅值。

实验结果

研究问题

  • RQ1DeepONet 是否能够学习将上游不稳定扰动映射到高速边界层中的下游扰动场的算子?
  • RQ2对于未见扰动频率和相位,DeepONet 能否多准确地预测下游场?
  • RQ3DeepONet 是否能够解决从下游壁压测量中推断上游扰动的逆问题?
  • RQ4在从同一控制方程导出的不同场(速度、压力、温度)进行预测时,DeepONet 的表现如何比较?
  • RQ5结合正向和逆向 DeepONet 进行基于壁压的扰动重建的数据同化工作流是什么?

主要发现

  • DeepONets 能够再现 PSE 预测的下游扰动场,在验证模态上的相对误差低于 5%。
  • 正向情形在未见频率下显示出壁法向剖面、波长、相位和振幅的准确恢复。
  • 逆向情形从壁压测量中成功重建下游输入扰动,建立一个数据同化循环。
  • 训练需要大量数据生成,但在对新输入求解 PSE 相比之下,评估速度极快。
  • 特征扩展与损失加权有助于缓解由于模态幅值差异引起的训练挑战,提升收敛性和准确性。
  • 每次 DeepONet 评估的时间比运行新的 PSE 仿真快出数量级,使得快速多查询评估成为可能。

更好的研究,从现在开始

从阅读论文到最终审阅,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。