One-Step Flow Matching for Generative Modeling of Path-Dependent Physical Fields
作者: Yijing Zhou, Jasmin Jelovica
分类: cs.LG, cs.CE, physics.comp-ph
发布日期: 2026-06-22
备注: 25 pages
💡 一句话要点
提出基于流匹配的生成模型以高效模拟路径依赖物理场
🎯 匹配领域: 支柱二:RL算法与架构 (RL & Architecture)
关键词: 路径依赖模型 生成模型 流匹配 变换器 物理仿真 高效计算 非高斯分布 有限元分析
📋 核心要点
- 现有的路径依赖物理场仿真方法计算成本高,尤其是使用传统的有限元分析时,效率低下。
- 本文提出了一种基于变换器的流匹配模型,通过在潜在空间中直接生成应力场,简化了仿真过程。
- 实验结果显示,该模型在有限数据集上仍能生成高分辨率的路径依赖场,计算效率显著提升。
📝 摘要(中文)
复杂几何体的路径依赖本构模型物理仿真面临巨大的计算成本。近年来,生成AI模型在图像和视频合成任务中的成功为改善仿真提供了希望。尽管基于U-Net的去噪扩散概率模型(DDPMs)已被用于弹性应力场生成,但通常需要数百个采样步骤,且生成路径依赖(如塑性)应力场的应用仍然有限。本文提出了一种基于变换器骨干的流匹配(FM)模型,用于高分辨率路径依赖应力场生成,能够在随机加载-卸载路径和几何体下直接生成应力场。我们的模型在变分自编码器(VAE)的潜在空间内运行,设计了非高斯源分布以减少训练过程中的条件传输路径交叉,从而实现一步生成满意样本,且无需依赖蒸馏。实验结果表明,即使在有限的训练数据集上,我们的模型也能准确生成高分辨率路径依赖场,计算效率比有限元分析高出6到7倍,在消费级GPU上速度提升近两个数量级。
🔬 方法详解
问题定义:本文旨在解决复杂几何体的路径依赖物理场仿真中计算成本高的问题。现有方法如有限元分析效率低下,尤其在处理塑性应力场时,通常需要大量的采样步骤。
核心思路:提出的流匹配模型基于变换器架构,利用潜在空间生成应力场,将仿真视为视频合成任务,从而实现高效生成。
技术框架:该模型在变分自编码器(VAE)的潜在空间中运行,包含流匹配模块、非高斯源分布设计和辅助网络,能够直接生成所有时间步的应力场。
关键创新:最重要的创新在于设计了非高斯源分布以减少条件传输路径的交叉,允许模型在一步内生成高质量样本,避免了传统方法的蒸馏过程。
关键设计:模型中引入了令牌级加载嵌入和两个辅助网络,以增强路径依赖仿真的性能,损失函数和网络结构经过精心设计以适应流匹配的需求。
🖼️ 关键图片
📊 实验亮点
实验结果表明,所提模型在有限训练数据集上能够准确生成高分辨率的路径依赖应力场,计算效率比传统有限元分析高出6到7倍,在消费级GPU上速度提升近两个数量级,展示了其优越的性能。
🎯 应用场景
该研究的潜在应用领域包括工程仿真、材料科学和结构分析等,能够显著提高路径依赖物理场的仿真效率,降低计算成本,推动相关领域的研究与应用发展。未来,该模型可能在实时仿真和优化设计中发挥重要作用。
📄 摘要(原文)
Physical simulations for intricate geometries with path-dependent constitutive models face difficulties due to the enormous computational cost they require. Recently, the emergence of generative AI models, which succeed in image and video synthesis tasks, has provided a promise to further improve simulations. Although U-Net-based denoising diffusion probabilistic models (DDPMs) have been adopted for elastic stress field generation, they typically require hundreds of sampling steps, and applications of generative models to path-dependent, e.g. plastic, stress fields remain very limited. In this work, we propose a novel flow matching (FM) model based on a transformer backbone for high-resolution path-dependent stress field generation with stochastic loading-unloading paths and geometry. The proposed model operates within the latent space of a variational autoencoder (VAE) and formulates the simulation of plastic fields as a video synthesis task, directly generating the stress fields across all time steps. Meanwhile, we design a non-Gaussian source distribution for flow matching, such that crossings among conditional transport paths are reduced during training. This enables our model to generate satisfactory samples in one step without relying on distillation. In addition, we introduce token-level loading embeddings and two auxiliary networks to further enhance the model performance in path-dependent simulation. The results demonstrate that, even with a limited training dataset, our model can accurately generate high-resolution path-dependent fields. It is much more computationally efficient than finite element analysis, providing a speedup of 6 to 7 times over FEM on CPUs and approximately two orders of magnitude speedup on consumer-grade GPUs.