含能材料2024，Vol.32Issue(4)：408-421,14.DOI:10.11943/CJEM2023226
人工智能辅助含能分子设计的应用与展望
Applications and Prospects of AI-assisted Design of Energetic Molecules
摘要
Abstract
The explore of energetic molecules faces multiple challenges,and the traditional design method are inefficient.The emergence of computer-aided molecular design has changed the research and development model.This review provides an over-view of the development of energetic molecular design and introduces the current research status of computer-aided energetic molecular design.By summarizing the latest advancements in Artificial Intelligence(AI)technology across various design as-pects,including performance prediction,molecular generation,retrosynthetic reaction prediction,and reaction condition pre-diction,we discussed the existing gap between the current approaches in energetic molecular design and other materials design methods.By thinking about the causes of the gap,we present an outlook on the future developmental directions of AI-assisted en-ergetic molecular design.Research indicates that AI has already been applied in property prediction and molecular generation of energetic molecular design,but requires further exploration in retrosynthetic reaction prediction,and reaction conditions predic-tion.AI-assisted design of energetic molecules holds broad promising application prospects.Data enhancement,transfer learning and high-throughput computing are expected to solve the problem of weak data of energetic molecules.Enhancing AI-assisted prediction of synthesis routes and reaction conditions for energetic molecules shows promise for achieving the automatic molecu-lar design via whole process of"design→evaluation→preparation→verification".AI-assisted energetic molecular design provides new possibilities for improving the level of energetic molecular design and helps to improve the efficiency of energetic molecule research and development.关键词
含能分子/分子设计/人工智能/机器学习/定量构效关系Key words
energetic molecule/molecular design/artificial intelligence/machine learning/quantitative structure-property rela-tionships分类
军事科技引用本文复制引用
刘锐,刘建,唐岳川,张朝阳,黄静,黄鑫..人工智能辅助含能分子设计的应用与展望[J].含能材料,2024,32(4):408-421,14.基金项目
国家自然科学基金(22173086,22203081,22305234) National Natural Science Foundation of China(Nos.22173086,22203081,22305234) （22173086,22203081,22305234）
