.png){kind=icon}
返回
Efficient neural network training method for unsteady flow field prediction based on data pool
.png){kind=icon}
OA
AI
DOI：10.1080/19942060.2025.2512958.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
In recent years, neural network technology has made significant progress in the field of unsteady flow field prediction, leading to the development of many innovative methods. However, deep learning-based unsteady flow field prediction techniques typically rely on autoregressive models, which inevitably face the issue of error accumulation. Existing solutions often suffer from challenges such as complex hyperparameter configurations and reduced training efficiency. To address these issues, this study makes the following contributions: (1) A novel training method is proposed to enhance model convergence by refining the training strategies employed. This approach achieves improved performance without necessitating complex hyperparameter configurations. (2) An innovative curriculum learning-based timestep reset strategy is introduced. This strategy further improves the convergence of neural networks and enhances prediction accuracy. A detailed comparative study was conducted on different training methods within the architecture of convolutional neural networks. Experimental results show that the proposed training strategy significantly improves prediction accuracy, with an improvement of up to an order of magnitude. Moreover, even when the training set spans only 600 timesteps, the model remains stable when predicting up to 9000 timesteps. Finally, our method also demonstrates high efficiency in terms of training time.
Keyword:
Neural network
unsteady flow field prediction
attention U-net
data pool
期刊
IF:
5.4
论文数:
1.4K
被引数:
3.9K
机构
引用论文
Nakamura, T., Fukami, K., Hasegawa, K., Nabae, Y., & Fukagata, K. (2021). Convolutional neural network and long short-term memory based reduced order surrogate for minimal turbulent channel flow. Physics of Fluids, 33(2), 023601. https://doi.org/10.1063/5.0039845 (Open in a new window)Web of Science ®(Open in a new window)Google Scholar中村, T., 深見, K., 長谷川, K., 半井, Y., & 深方, K. (2021). 基于卷积神经网络和长短时记忆网络的极小湍流通道流降阶代理模型. 流体物理学, 33(2), 023601. https://doi.org/10.1063/5.0039845 (在新窗口中打开)Web of Science ®(在新窗口中打开)Google Scholar.png)
IF0
PREAI
分享收藏Simpson, R. L. (1981). A review of some phenomena in turbulent flow separation. Journal of Fluid Mechanics, 110, 1–20. https://doi.org/10.1017/S0022112081003033 (Open in a new window)Google Scholar辛普森, R. L. (1981). 湍流分离中某些现象的综述. 流体力学杂志, 110, 1–20. https://doi.org/10.1017/S0022112081003033 (在新窗口中打开)谷歌学术
IF0
PREAI
分享收藏Chen, X., Wang, Z., Deng, L., Yan, J., Gong, C., Yang, B., Wang, Q., Zhang, Q., Yang, L., Pang, Y., & Liu, J. (2024). Towards a new paradigm in intelligence-driven computational fluid dynamics simulations. Engineering Applications of Computational Fluid Mechanics, 18(1), 2407005. https://doi.org/10.1080/19942060.2024.2407005 (Open in a new window)Web of Science ®(Open in a new window)Google Scholar陈，X.，王，Z.，邓，L.，闫，J.，龚，C.，杨，B.，王，Q.，张，Q.，杨，L.，庞，Y.，& 刘，J. (2024)。迈向智能驱动的计算流体动力学模拟新范式。计算流体力学工程应用，18(1)，2407005。https://doi.org/10.1080/19942060.2024.2407005 (在新窗口打开)Web of Science ®(在新窗口打开)谷歌学术
IF0
PREAI
分享收藏Sun, L., & Wang, J. X. (2020). Physics-constrained bayesian neural network for fluid flow reconstruction with sparse and noisy data. Theoretical and Applied Mechanics Letters, 10(3), 161–169. https://doi.org/10.1016/j.taml.2020.01.031 (Open in a new window)Web of Science ®(Open in a new window)Google ScholarSun, L., & Wang, J. X. (2020). 物理约束贝叶斯神经网络用于稀疏和噪声数据下的流体流场重构. 理论与实验力学学报, 10(3), 161–169. https://doi.org/10.1016/j.taml.2020.01.031 (在新窗口打开)Web of Science ®(在新窗口打开)Google Scholar
IF0
PREAI
分享收藏Geneva, N., & Zabaras, N. (2022). Transformers for modeling physical systems. Neural Networks, 146, 272–289. https://doi.org/10.1016/j.neunet.2021.11.022 (Open in a new window)PubMed (Open in a new window)Web of Science ®(Open in a new window)Google ScholarGeneva, N., & Zabaras, N. (2022). 适用于物理系统建模的转换器。神经网络, 146, 272–289. https://doi.org/10.1016/j.neunet.2021.11.022 (在新窗口中打开)PubMed (在新窗口中打开)Web of Science ®(在新窗口中打开)Google Scholar
IF0
PREAI
分享收藏Maulik, R., Fukami, K., Ramachandra, N., Fukagata, K., & Taira, K. (2020). Probabilistic neural networks for fluid flow surrogate modeling and data recovery. Physical Review Fluids, 5(10), 104401. https://doi.org/10.1103/PhysRevFluids.5.104401 (Open in a new window)Web of Science ®(Open in a new window)Google ScholarMaulik, R., Fukami, K., Ramachandra, N., Fukagata, K., & Taira, K. (2020). 概率神经网络用于流体流动代理建模与数据恢复. 物理评论流体, 5(10), 104401. https://doi.org/10.1103/PhysRevFluids.5.104401 (在新窗口中打开)Web of Science ®(在新窗口中打开)Google Scholar
IF0
PREAI
分享收藏Linot, A. J., & Graham, M. D. (2023). Dynamics of a data driven low dimensional model of turbulent minimal Couette flow. Journal of Fluid Mechanics, 973, A42. https://doi.org/10.1017/jfm.2023.720 (Open in a new window)Web of Science ®(Open in a new window)Google Scholar林诺特，A. J.，& 格拉汉姆，M. D. (2023)。数据驱动低维模型下湍流最小库埃特流的动力学。流体力学杂志，973，A42。https://doi.org/10.1017/jfm.2023.720 (在新窗口中打开)Web of Science ®(在新窗口中打开)谷歌学术
IF0
PREAI
分享收藏Kingma, D. P., and J. Ba. 2014. “Adam: A Method for Stochastic Optimization.” arXiv preprint arXiv:1412.6980. https://doi.org/10.48550/arXiv.1412.6980. (Open in a new window)Google ScholarKingma, D. P., and J. Ba. 2014. “Adam: 一种随机优化方法。” arXiv预印本 arXiv:1412.6980. https://doi.org/10.48550/arXiv.1412.6980. (在新窗口中打开)谷歌学术
IF0
PREAI
分享收藏Solera-Rico, A., Sanmiguel Vila, C., Gómez-López, M., Wang, Y., Almashjary, A., Dawson, S. T. M., & Vinuesa, R. (2024). β-variational auto encoders and transformers for reduced-order modelling of fluid flows. Nature Communications, 15(1), 1361. https://doi.org/10.1038/s41467-024-45578-4 (Open in a new window)PubMed(Open in a new window)Google ScholarSolera-Rico, A., Sanmiguel Vila, C., Gómez-López, M., Wang, Y., Almashjary, A., Dawson, S. T. M., & Vinuesa, R. (2024). β-变分自编码器和变换器用于流体流动的降阶建模。自然通讯，15（1），1361。https://doi.org/10.1038/s41467-024-45578-4 (在新窗口中打开)PubMed(在新窗口中打开)谷歌学术
IF0
PREAI
分享收藏Li, T., Zou, S., Chang, X., Zhang, L., & Deng, X. (2024). Predicting unsteady incompressible fluid dynamics with finite volume informed neural network. Physics of Fluids, 36(4), 043601. https://doi.org/10.1063/5.0197425 (Open in a new window)Web of Science ®(Open in a new window)Google Scholar李, T., 邹, S., 常, X., 张, L., & 邓, X. (2024). 使用有限体积信息神经网络预测非定常不可压缩流体动力学. 流体物理学, 36(4), 043601. https://doi.org/10.1063/5.0197425 (在新窗口打开)Web of Science ®(在新窗口打开)谷歌学术
IF0
PREAI
分享收藏
查看更多内容