数据驱动的材料智能设计平台研究与应用  

Research and Application of a Data-Driven Intelligent Design Platform for Materials

作　　者：王宗国[1,2] 万萌 陈子逸 李凯 王晓光[1] 刘淼 孟胜 王彦棡[1,2] WANG Zongguo;WAN Meng;CHEN Ziyi;LI Kai;WANG Xiaoguang;LIU Miao;MENG Sheng;WANG Yangang(Computer Network Information Center,Chinese Academy of Sciences,Beijing 100083,China;University of Chinese Academy of Sciences,Beijing 100049,China;Institute of Physics,Chinese Academy of Sciences,Beijing 100190,China)

机构地区：[1]中国科学院计算机网络信息中心,北京100083 [2]中国科学院大学,北京100049 [3]中国科学院物理研究所,北京100190

出　　处：《数据与计算发展前沿》2023年第2期86-96,共11页Frontiers of Data & Computing

基　　金：国家重点研发计划(2020YFB0204802);中国科学院前沿科学重点研究计划(ZDBS-LY-7025);中国科学院网信专项(CAS-WX2021PY-0102);中国科学院青年创新促进会。

摘　　要：【目的】新科研范式下,基于大数据的人工智能技术为加速新材料设计与发现提供了新的方法与视角,为材料领域研究人员提供了一个可用的材料智能设计平台,对于新材料的发现与性能优化具有重要意义。【方法】本文提出一种基于数据驱动的材料智能设计平台的整体架构,阐述应用平台开展新材料设计和优化的关键技术及相关工具,并给出该平台在材料领域的应用案例。【结果】材料智能设计平台及其应用加快了新材料设计与性能优化的进程,同时也为科研人员提供了一种交互式、插件式的开发环境。【局限】材料领域数据的多源异构、样本小、含噪声且数据关系复杂等特点给模型训练效果产生一定的影响,未来希望在数据规范和小样本训练方面进行更多探索。【结论】本文所提出的材料设计平台为材料领域实现科研范式变革提供了理论依据和示范作用。[Objective]Under the new scientific paradigm,artificial intelligence(AI)based on big data has provided new methods and perspectives for accelerating new material design and discovery.Pro-viding a useful intelligent design platform for material researchers has significant implications for im-proving the discovery efficiency and performance optimization of new materials.[Methods]This article proposes an overall architecture of a data-driven intelligent design platform for materials,elaborating on the key technologies and related tools used to develop and optimize new materials on the platform.Additionally,the article provides an application case for the platform in material science.[Results]The material intelligent design platform and its application have accelerated the process of new material design and performance optimization,while also providing researchers with an interactive and plugin-based development environment.[Limitations]The characteristics of the material data,such as multiple heterogeneous sources,small sample sizes,and complex relationships,have a particular impact on the training effect of the models.In the future,more exploration will be done on data standardization and small sample training.[Conclusions]The material design platform proposed in this article provides a theoretical basis and demonstration for the transformation of the scientific research paradigm in material science.

关 键 词：数据驱动 人工智能模型 特征计算 材料设计平台 科研范式 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TB30[自动化与计算机技术—控制科学与工程]