arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2301.07541physics.flu-dyncs.CVcs.LGnlin.CDphysics.data-an
收藏

用于推断旋转湍流中速度分量的生成对抗网络

Generative Adversarial Networks to infer velocity components in rotating turbulent flows

University of Rome “Tor Vergata”(罗马第二大学);INFN(意大利国家核物理研究所)

Tianyi Li, Michele Buzzicotti, Luca Biferale, Fabio Bonaccorso

AI总结:

针对旋转湍流二维快照的速度分量推断问题,对比测试EPOD、CNN和GAN的重建能力,发现GAN在弱相关场景下仍能实现统计意义上的流场重建,性能更优。

摘要:

本文研究了旋转湍流二维快照的推断问题。我们对线性扩展本征正交分解(Extended Proper Orthogonal Decomposition,EPOD)方法、非线性卷积神经网络(Convolutional Neural Network,CNN)以及生成对抗网络(Generative Adversarial Network,GAN)的逐点重建能力和统计重建能力开展了系统的定量基准测试。我们聚焦于通过测量一个速度分量来推断另一个速度分量的重要任务,共研究两种情形:(I)两个分量都位于与旋转轴正交的平面内;(II)两个分量中有一个与旋转轴平行。研究表明,EPOD方法仅在两个分量强相关的前一种情形下表现良好,而CNN和GAN在逐点重建和统计重建方面始终优于EPOD。对于情形(II),当输入和输出数据弱相关时,所有方法都无法忠实地重建逐点信息。在这种情况下,只有GAN能够从统计意义上重建流场。分析过程既使用了基于预测结果与真实值之间$L_2$空间距离的标准验证工具,也采用了更精细的基于小波分解的多尺度分析。统计验证则基于概率密度函数之间的标准Jensen-Shannon散度、谱特性以及多尺度平坦度。

英文摘要:

Inference problems for two-dimensional snapshots of rotating turbulent flows are studied. We perform a systematic quantitative benchmark of point-wise and statistical reconstruction capabilities of the linear Extended Proper Orthogonal Decomposition (EPOD) method, a non-linear Convolutional Neural Network (CNN) and a Generative Adversarial Network (GAN). We attack the important task of inferring one velocity component out of the measurement of a second one, and two cases are studied: (I) both components lay in the plane orthogonal to the rotation axis and (II) one of the two is parallel to the rotation axis. We show that EPOD method works well only for the former case where both components are strongly correlated, while CNN and GAN always outperform EPOD both concerning point-wise and statistical reconstructions. For case (II), when the input and output data are weakly correlated, all methods fail to reconstruct faithfully the point-wise information. In this case, only GAN is able to reconstruct the field in a statistical sense. The analysis is performed using both standard validation tools based on $L_2$ spatial distance between the prediction and the ground truth and more sophisticated multi-scale analysis using wavelet decomposition. Statistical validation is based on standard Jensen-Shannon divergence between the probability density functions, spectral properties and multi-scale flatness.

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。

↑