发表机构
Swiss Federal Institute of Technology in Lausanne (EPFL)（洛桑联邦理工学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出DiffGeo框架，结合潜在空间模型与扩散采样器，在极端数据稀缺下实现气动外形设计空间探索，经多案例验证其性能优于基线方法，可提升设计效率与多样性。
AI 中文摘要
我们提出DiffGeo，一种用于极端数据稀缺场景下气动设计空间探索的基于潜在空间扩散的生成框架。DiffGeo结合了用于自动外形参数化的学习型潜在空间模型，以及扩散采样器，可直接生成新颖、几何有效且可控的设计。我们通过一系列案例研究验证该方法：（i）二维翼型生成基准测试，在有限数据下，从样本质量、多样性和约束遵守度方面，将DiffGeo的潜在扩散模型与基于GAN和VAE的基线进行对比；（ii）集成到基于代理的优化流程中，DiffGeo的条件采样生成任务感知的翼型数据，可提升代理建模和优化性能；（iii）扩展到三维涡轮机械叶片原型设计，DiffGeo从少量参考设计生成逼真且高性能的叶片几何结构。在所有研究中，DiffGeo实现了高质量且多样的外形生成，所需数据量至少比替代方法少一个数量级，将几何表示与设计目标解耦以实现灵活复用，并通过基于能量的条件设置无缝融入复杂设计约束。这些能力表明DiffGeo有潜力通过自动化设计空间探索来增强早期设计——提升效率、扩展设计多样性，并通过可控引导嵌入工程知识。
英文摘要
We propose DiffGeo, a latent space diffusion-based generative framework for aerodynamic design space exploration under extreme data scarcity. DiffGeo combines a learned latent space model for automatic shape parameterization, with a diffusion sampler to directly generate novel, geometry-valid and controllable designs. We validate the approach on a series of case studies: (i) a 2D airfoil generation benchmark, where DiffGeo's latent diffusion model is compared against GAN- and VAE-based baselines in terms of sample quality, diversity and constraint adherence under limited data; (ii) integration into a surrogate-based optimization pipeline, where DiffGeo's conditional sampling produces task-informed airfoil data that improve both surrogate modeling and optimization performance; and (iii) extension to 3D turbomachinery blade prototyping, where DiffGeo generates realistic and high-performance blade geometries from a small set of reference designs. Throughout these investigations, DiffGeo achieves high-quality and diverse shape generation with at least an order of magnitude less data than alternatives, decouples geometry representation from design targets for flexible reuse, and seamlessly incorporates complex design constraints via energy-based conditioning. These capabilities demonstrate DiffGeo's potential to enhance early-stage design by automating design space exploration--improving efficiency, expanding design diversity and embedding engineering knowledge through controllable guidance.
CommentsThis is the authors' postprint of the AIAA Journal article https://doi.org/10.2514/1.J066320. Data, code and agentic skill demo: https://github.com/kfxw/DiffGeo
Journal refAIAA Journal, 2026
