Scholarlink

<p>Efficiently and accurately forecasting the long-term behavior of turbulence remains a major challenge for both conventional numerical simulations and machine learning techniques. Recently, neural operators have emerged as a promising solution. The U-Net enhanced implicit Fourier neural operator (IU-FNO) has proven effective in generating stable long-term predictions for three-dimensional incompressible turbulence. This study extends IU-FNO to three-dimensional chemically reacting compressible turbulence. Numerical experiments indicate that IU-FNO predicts flow dynamics much faster than the dynamic Smagorinsky model (DSM) employed in large eddy simulations. In terms of prediction accuracy, the IU-FNO framework exhibits superior performance compared to the DSM in capturing the energy spectra of velocity, temperature, and density, the probability density functions of vorticity and velocity increments, and the instantaneous spatial structures of temperature. Consequently, IU-FNO represents a highly effective approach for modeling chemically reacting compressible turbulence.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of three-dimensional chemically reacting compressible turbulence based on implicit U-Net enhanced Fourier neural operator

  • 期刊论文

  • 发表日期：

    2026年4月20日

查看全文
  • Zhiyao Zhang,
  • Zhijie Li,
  • Yunpeng Wang,
  • Huiyu Yang,
  • Wenhui Peng,
  • Jian Teng,
  • Jianchun Wang

摘要

Efficiently and accurately forecasting the long-term behavior of turbulence remains a major challenge for both conventional numerical simulations and machine learning techniques. Recently, neural operators have emerged as a promising solution. The U-Net enhanced implicit Fourier neural operator (IU-FNO) has proven effective in generating stable long-term predictions for three-dimensional incompressible turbulence. This study extends IU-FNO to three-dimensional chemically reacting compressible turbulence. Numerical experiments indicate that IU-FNO predicts flow dynamics much faster than the dynamic Smagorinsky model (DSM) employed in large eddy simulations. In terms of prediction accuracy, the IU-FNO framework exhibits superior performance compared to the DSM in capturing the energy spectra of velocity, temperature, and density, the probability density functions of vorticity and velocity increments, and the instantaneous spatial structures of temperature. Consequently, IU-FNO represents a highly effective approach for modeling chemically reacting compressible turbulence.