Wang Zhuo, Yang Wenhua, Xiang Linyan, Wang Xiao, Zhao Yingjie, Xiao Yaohong, Liu Pengwei, Liu Yucheng, Banu Mihaela, Zikanov Oleg, Chen Lei
Department of Mechanical Engineering, University of Michigan-Dearborn, Dearborn, MI 48128, USA.
Department of Mechanical Engineering, Mississippi State University, Starkville, MS 39762, USA.
Patterns (N Y). 2022 Apr 21;3(6):100494. doi: 10.1016/j.patter.2022.100494. eCollection 2022 Jun 10.
There is a compelling need for the regression capability of mapping the initial field and applied conditions to the evolved field, e.g., given current flow field and fluid properties predicting next-step flow field. Such a capability can provide a maximum to full substitute of a physics-based model, enabling fast simulation of various field evolvements. We propose a conceptually simple, lightweight, but powerful multi-input convolutional network (ConvNet), yNet, that merges multi-input signals by manipulating high-level encodings of field/image input. yNet can significantly reduce the model size compared with its ConvNet counterpart (e.g., to only one-tenth for main architecture of 38-layer depth) and is as much as six orders of magnitude faster than a physics-based model. yNet is applied for data-driven modeling of fluid dynamics, porosity evolution in sintering, stress field development, and grain growth. It consistently shows great extrapolative prediction beyond training datasets in terms of temporal ranges, spatial domains, and geometrical shapes.
迫切需要具备将初始场和应用条件映射到演化场的回归能力，例如，给定当前流场和流体属性来预测下一步的流场。这样的能力可以最大程度地完全替代基于物理的模型，从而能够快速模拟各种场的演化。我们提出了一种概念上简单、轻量级但功能强大的多输入卷积网络（ConvNet），即yNet，它通过操纵场/图像输入的高级编码来合并多输入信号。与对应的卷积网络相比，yNet可以显著减小模型大小（例如，对于38层深度的主要架构，仅为其十分之一），并且比基于物理的模型快多达六个数量级。yNet应用于流体动力学、烧结过程中的孔隙率演化、应力场发展和晶粒生长的数据驱动建模。在时间范围、空间域和几何形状方面，它始终在训练数据集之外表现出出色的外推预测能力。
