AI Insight
Researchers developed a multi-stride neural operator that improves long-term prediction of three-dimensional turbulent flows by combining multiple prediction timescales. The method uses specialized transformer-based predictors at different stride lengths and a fusion module that interpolates and corrects intermediate velocity fields, allowing the system to decouple autoregressive steps from output time steps. Testing on turbulent channel flow and homogeneous isotropic turbulence showed the approach reduced statistical prediction errors by over 98% compared to fixed-stride methods while generating fine-timestep trajectories up to 3.69 times faster.
Why it matters
Accurate turbulence prediction is critical for applications in weather forecasting, climate modeling, aircraft design, and combustion systems. This method offers a pathway to significantly reduce computational costs of turbulence simulations while maintaining statistical accuracy over long time periods, potentially enabling more detailed and longer-duration predictions in fluid dynamics applications.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
Abstract: Long-time turbulence prediction at fine time steps is costly and prone to error accumulation, while accurate short-term fields do not ensure accurate long-time statistics. We propose a multi-stride neural operator for three-dimensional turbulence, combining implicit factorized Transformer (IFF) predictors with a factorized-attention cross-stride fusion (FA-CSF) module. All predictors take the same current velocity field as input. FA-CSF reuses their outputs and latent features to generate intermediate velocity fields through interpolation and learned correction. The corrected largest-stride field drives the next autoregressive step, decoupling the autoregressive step from the output time step. Stride-conditioned IFF (SC-IFF) predictors are first trained at power-of-two strides using three conditioning strategies, then frozen while FA-CSF is trained on long rollouts with field-data and statistical losses. Tests on turbulent channel flow and forced homogeneous isotropic turbulence (HIT) show improved long-time predictions and statistics. At a sampling interval of 10 direct numerical simulation (DNS) time steps, fixed-stride IFF yields relative L2 errors above 100% in channel-flow root-mean-square (RMS) velocity fluctuation profiles and the Reynolds shear stress profile, and in the temporal evolution of velocity and vorticity RMS values and the time-averaged energy spectrum in HIT. SC-IFF with independent stride-specific predictors and FA-CSF keeps statistics close to filtered DNS data over the tested prediction times. Relative to fixed-stride IFF, it reduces errors in the channel-flow profiles by more than 98% and in the HIT spectrum by more than 99%. Multi-stride prediction with FA-CSF generates trajectories at fine time steps 3.69 times faster than fixed-stride IFF autoregression and 2.41 times faster than hierarchical multi-stride prediction.
Source: Long-time turbulence prediction at fine time steps using multi-stride neural operators