Predicting unavailable parameters from existing velocity fields of turbulent flows using a GAN-based model

TL;DR

This paper introduces M-GAN, a deep learning model that predicts missing parameters like temperature and pressure from available velocity data in turbulent flows, demonstrating high accuracy across different flow conditions.

Contribution

The study presents a novel GAN-based framework with label information generation for predicting unavailable flow parameters from existing velocity fields.

Findings

M-GAN accurately predicts temperature from velocity data in Rayleigh-Bénard flow.

The model successfully estimates pressure and velocity fields in turbulent channel flows.

Predictions closely match DNS data even at higher Reynolds numbers.

Abstract

In this study, an efficient deep-learning model is developed to predict unavailable parameters, e.g., streamwise velocity, temperature, and pressure from available velocity components. This model, termed mapping generative adversarial network (M-GAN), consists of a label information generator (LIG) and an enhanced super-resolution generative adversarial network (ESRGAN). LIG can generate label information helping the model to predict different parameters. The GAN-based model receives the label information from LIG and existing velocity data to generate the unavailable parameters. Two-dimensional (2D) Rayleigh-B{\'e}nard flow and turbulent channel flow are used to evaluate the performance of M-GAN. Firstly, M-GAN is trained and evaluated by 2D direct numerical simulation (DNS) data of a Rayleigh-B{\'e}nard flow. From the results, it can be shown that M-GAN can predict temperature distribution from the two-dimensional velocities. Furthermore, DNS data of turbulent channel flow at two different friction Reynolds numbers $Re_\tau$ = 180 and 550 are applied simultaneously to train the M-GAN and examine its predicting ability for the pressure fields and the streamwise velocity from the other two velocity components. The instantaneous and statistical results of the predicted data agree well with the DNS data, even for the flow at $Re_\tau$ = 395, indicating that M-GAN can be trained to learn the mapping function of the unknown fields with good interpolation capability.