无机材料学报 ›› 2019, Vol. 34 ›› Issue (1): 27-36.DOI: 10.15541/jim20180214

所属专题: MAX相和MXene材料 副主编黄庆研究员专辑

• 综述 • 上一篇    下一篇

自然启发算法库构建设想及其在新材料研发中的意义

都时禹1, 张一鸣1, 罗侃1,2, 黄庆1   

  1. 1. 中国科学院 宁波材料技术与工程研究所, 核能材料工程实验室(筹), 宁波 315201;
    2. 华东理工大学 化工学院, 上海200237
  • 收稿日期:2018-05-08 修回日期:2018-07-13 出版日期:2019-01-21 发布日期:2018-12-17
  • 基金资助:
    国家重点研发计划(2016YFB0700100);中国科学院前沿科学重点研究计划(QYZDB-SSW-JSC037);中国科学院王宽诚率先人才计划卢嘉锡国际团队项目(rczx0800);中国科学院创新交叉团队(关键核能技术交叉团队);National Key Research and Development Program of China (2016YFB0700100);Key Research Program of Frontier Sciences, Chinese Academy of Sciences (QYZDB-SSW-JSC037);K. C. Wong Education Foundation (rczx0800);Key Technology of Nuclear Energy, 2014, CAS Interdisciplinary Innovation Team

Design of the Nature-inspired Algorithms Library and Its Significance for New Materials Research and Development

DU Shi-Yu1, ZHANG Yi-Ming1, LUO Kan1,2, HUANG Qing1   

  1. 1. Engineering Laboratory of Nuclear Energy Materials, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo 315201, China;
    2. School of Chemical Engineering, East China University of Science and Technology, Shanghai 200237, China
  • Received:2018-05-08 Revised:2018-07-13 Online:2019-01-21 Published:2018-12-17

摘要:

材料基因组工程技术是运用人工智能手段实现新材料按需设计的关键技术, 其中尤为重要的是创新智能算法的开发和应用。本文在总结、分析已有自然启发算法的基础上, 提出建立自然启发算法库(Nature-inspired Algorithms Library, NIAL)的设想; 明确了从不同学科取得算法启发并高通量产生新算法的基本思路; 详细阐述了构建该算法库的基本流程, 并剖析建立自然启发算法库平台的若干优势和特点。最后, 展望了自然启发算法库在新材料研发中的应用模式, 希望借此提升人工智能在材料基因组工程领域的应用水平。

 

关键词: 自然启发算法, 材料基因组工程, 人工智能, 综述

Abstract:

The technique for Materials Genetic Initiative (MGI) is the key tool for realizing the demand-oriented design of new materials assisted by the artificial intelligence (AI). Accordingly, the development and application of innovative intelligence algorithms are particularly important. Based on the generalization and analyses of the existing nature-inspired algorithms, this work aims at outlining the suggestion to build the nature-inspired algorithms library (NIAL). The potential route in which inspirations are obtained from varieties of disciplines, was used to produce new algorithms in high-throughput ways is introduced. The general procedure for building algorithm library is elaborated, while its advantages and characteristics are anatomized. Finally, the potential of NIAL in new materials development has been envisioned to enhance the standard for the application of AI including MGI.

Key words: Nature-Inspired Algorithms Library, Materials Genetic Initiative, artificial intelligence, review

中图分类号: