Long-horizon prediction of three-dimensional wall-bounded turbulence with CTA-Swin-UNet and resolvent analysis

2026-05-18

AI Summary

This study addresses the challenges of rapid error accumulation and high computational cost in long-horizon predictions of three-dimensional wall-bounded turbulence. To this end, the authors propose a hybrid machine learning framework that integrates a channel-temporal attention (CTA) mechanism into Swin-UNet and incorporates multi-timescale fusion correction (MTFC) to enhance prediction stability. Furthermore, the method reconstructs the full three-dimensional flow field by leveraging response mode analysis combined with spectral linear stochastic estimation (SLSE). The model achieves stable autoregressive rollouts up to 150 steps, which extends to 300 steps when augmented with MTFC—significantly outperforming baseline approaches such as LSTM and Fourier Neural Operator (FNO)—while accurately capturing key features of three-dimensional turbulent structures and their energy spectra.

Abstract

Long-horizon prediction of three-dimensional (3D) wall-bounded turbulence with machine-learning methods remains a challenging task, due to the rapid accumulation of autoregressive errors and the substantially computational cost. To address these challenges, we present a hybrid machine-learning framework, in which a channel-time-attention Swin-UNet (CTA-Swin-UNet) and a multi-time-scale fusion correction (MTFC) strategy are developed to predict the turbulent flow fields in a wall-parallel plane, with affordable computational cost. Then, 3D flow fields are reconstructed via a resolvent-based spectral linear stochastic estimation (SLSE), rooting from the predicted planar flow. Results show that the CTA-Swin-UNet outperforms the baseline models (LSTM, FNO and traditional Swin-UNet) in both single-step prediction and autoregressive rollouts, indicating the effectiveness of introducing the CTA module into the Swin-UNet architecture. At the same temporal interval, the CTA-Swin-UNet remains stable for approximately 150 rollout steps, while the baseline models fail within 20 to 50 rollout steps. After introducing the MTFC strategy, a longer horizon upto 300 steps is achieved. Using the resolvent-based SLSE reconstruction further recovers the 3D flow structures and energy spectral distributions from the predicted planar inputs, which demonstrates that the proposed framework provides an effective and computationally efficient approach for long-horizon autoregressive prediction of 3D wall-bounded turbulence.

Problem

• long-horizon prediction
• 3D wall-bounded turbulence
• autoregressive error
• computational cost
• turbulent flow prediction

Innovation

• CTA-Swin-UNet
• resolvent analysis
• long-horizon prediction
• multi-time-scale fusion correction
• wall-bounded turbulence