当前位置:
X-MOL 学术
›
arXiv.cs.GR
›
论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Latent Space Subdivision: Stable and Controllable Time Predictions for Fluid Flow
arXiv - CS - Graphics Pub Date : 2020-03-12 , DOI: arxiv-2003.08723 Steffen Wiewel, Byungsoo Kim, Vinicius C. Azevedo, Barbara Solenthaler, Nils Thuerey
arXiv - CS - Graphics Pub Date : 2020-03-12 , DOI: arxiv-2003.08723 Steffen Wiewel, Byungsoo Kim, Vinicius C. Azevedo, Barbara Solenthaler, Nils Thuerey
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.
中文翻译:
潜在空间细分:流体流动的稳定可控时间预测
我们提出了一种端到端训练的神经网络架构,以稳健地预测具有高时间稳定性的流体流动的复杂动力学。我们专注于基于不可压缩 Navier-Stokes (NS) 方程的 2D 和 3D 单相烟雾模拟,这些方程与广泛的实际问题相关。为了实现对长期流序列的稳定预测,卷积神经网络 (CNN) 结合由堆叠的长短期记忆 (LSTM) 层组成的时间预测网络进行空间压缩训练。我们的核心贡献是一种新颖的潜在空间细分(LSS),用于将各个输入量分成编码的潜在空间域的各个部分。这允许在不干扰剩余潜在空间值的情况下有区别地改变编码量,从而最大化外部控制。通过选择性地覆盖部分预测的潜在空间点,我们提出的方法能够稳健地预测复杂物理问题的长期序列。此外,我们强调了由空间压缩网络执行的潜在空间创建的循环训练的好处。
更新日期:2020-03-20
中文翻译:
潜在空间细分:流体流动的稳定可控时间预测
我们提出了一种端到端训练的神经网络架构,以稳健地预测具有高时间稳定性的流体流动的复杂动力学。我们专注于基于不可压缩 Navier-Stokes (NS) 方程的 2D 和 3D 单相烟雾模拟,这些方程与广泛的实际问题相关。为了实现对长期流序列的稳定预测,卷积神经网络 (CNN) 结合由堆叠的长短期记忆 (LSTM) 层组成的时间预测网络进行空间压缩训练。我们的核心贡献是一种新颖的潜在空间细分(LSS),用于将各个输入量分成编码的潜在空间域的各个部分。这允许在不干扰剩余潜在空间值的情况下有区别地改变编码量,从而最大化外部控制。通过选择性地覆盖部分预测的潜在空间点,我们提出的方法能够稳健地预测复杂物理问题的长期序列。此外,我们强调了由空间压缩网络执行的潜在空间创建的循环训练的好处。