arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2409.19124cond-mat.mtrl-sciphysics.comp-ph
收藏

面向材料逆向设计的生成式深度学习

Generative deep learning for the inverse design of materials

Teng Long, Yixuan Zhang, Hongbin Zhang

AI总结:

本文综述了利用生成式深度学习进行材料逆向设计的方法,重点讨论晶体结构与微观结构的潜在空间构建、生成学习策略及性质约束,并指出当前在计算资源、数据兼容性和生成产率方面的挑战。

摘要:

除了利用机器学习进行材料性质的正向推断之外,应用于材料科学的生成式深度学习技术还允许对材料进行逆向设计,即以反向方式评估成分-工艺-(微观)结构-性质关系。在这篇综述中,我们聚焦于(微观)结构-性质映射,即晶体结构-本征性质和微观结构-非本征性质,并全面总结了如何执行生成式深度学习。详细讨论了三个关键要素,即晶体结构和微观结构的潜在空间构建、生成式学习方法以及性质约束。最后给出了一个展望,概述了现有方法在计算资源消耗、数据兼容性和生成产率方面面临的挑战。

英文摘要:

In addition to the forward inference of materials properties using machine learning, generative deep learning techniques applied on materials science allow the inverse design of materials, i.e., assessing the composition-processing-(micro-)structure-property relationships in a reversed way. In this review, we focus on the (micro-)structure-property mapping, i.e., crystal structure-intrinsic property and microstructure-extrinsic property, and summarize comprehensively how generative deep learning can be performed. Three key elements, i.e., the construction of latent spaces for both the crystal structures and microstructures, generative learning approaches, and property constraints, are discussed in detail. A perspective is given outlining the challenges of the existing methods in terms of computational resource consumption, data compatibility, and yield of generation.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑