Latent Space Subdivision: Stable and Controllable Time Predictions for Fluid Flow
This work addresses the challenge of stable and controllable time predictions for fluid flows, which is relevant for practical applications in physics simulations, but it appears incremental as it builds on existing CNN and LSTM methods with a novel subdivision technique.
The paper tackles the problem of predicting complex fluid flow dynamics with high temporal stability by proposing an end-to-end neural network architecture, achieving robust long-term predictions for single-phase smoke simulations in 2D and 3D based on the incompressible Navier-Stokes equations.
We propose an end-to-end trained neural networkarchitecture to robustly predict the complex dynamics of fluid flows with high temporal stability. We focus on single-phase smoke simulations in 2D and 3D based on the incompressible Navier-Stokes (NS) equations, which are relevant for a wide range of practical problems. To achieve stable predictions for long-term flow sequences, a convolutional neural network (CNN) is trained for spatial compression in combination with a temporal prediction network that consists of stacked Long Short-Term Memory (LSTM) layers. Our core contribution is a novel latent space subdivision (LSS) to separate the respective input quantities into individual parts of the encoded latent space domain. This allows to distinctively alter the encoded quantities without interfering with the remaining latent space values and hence maximizes external control. By selectively overwriting parts of the predicted latent space points, our proposed method is capable to robustly predict long-term sequences of complex physics problems. In addition, we highlight the benefits of a recurrent training on the latent space creation, which is performed by the spatial compression network.