机器学习在聚合物材料研究中的应用进展  

Review on application of machine learning in polymer material research

作　　者：王艺涵 贾丹[1,2] 詹胜鹏 徐宛星[1,2] 李健 段海涛 WANG Yi-han;JIA Dan;ZHAN Sheng-peng;XU Wan-xing;LI Jian;DUAN Hai-tao(State Key Laboratory of Special Surface Protection Materials and Application Technology,Wuhan Research Institute of Materials Protection,Wuhan 430030,China;Hubei Longzhong Laboratory,Xiangyang 441000,China)

机构地区：[1]武汉材料保护研究所,特种表面保护材料及应用技术国家重点实验室,湖北武汉430030 [2]湖北隆中实验室,湖北襄阳441000

出　　处：《高校化学工程学报》2023年第2期170-180,共11页Journal of Chemical Engineering of Chinese Universities

基　　金：国家自然科学基金(52175183);国家自然科学基金青年基金(51905385);武汉市科技局技术创新项目(2020010602012060)。

摘　　要：在大数据和人工智能相结合的现代科学研究的新形势下,聚合物材料性能的快速预测和新型聚合物材料的研发逐渐成为聚合物材料研究领域的关注焦点。将机器学习应用于聚合物材料研究领域,打破了传统试错法的局限性,通过数据直接建立材料特征与所需性能之间复杂的关系模型,解决聚合物组成成分和复杂结构等在其研究过程中带来的难题。论文介绍了机器学习在聚合物材料研究领域的常用方法及算法;总结了以宏观参量与微观参量为机器学习模型输入时,聚合物材料性能预测的研究进展;分析了基于机器学习的聚合物材料设计和新型聚合物材料研发的重要应用成果;讨论并提出当前基于机器学习的聚合物材料的研究热点与方向。With the development of big data and artificial intelligence,rapid prediction of polymer material properties and the development of new polymer materials gradually become the focus of polymer material research.The application of machine learning in polymer material studies breaks the limitation of traditional“trial-and-error”processes,establishes complex models between material characteristics and desired properties via data,and solves problems caused by polymer compositions and complex structures.This review introduces typical methods and algorithms of machine learning in polymer material research and summarizes research on predicting polymer material properties with macroscopic and microscopic parameters as the inputs.Moreover,important application achievements of machine learning-based polymer material design and new polymer material development were discussed.Finally,current research hot spots and directions based on machine learning were discussed and proposed.

关 键 词：机器学习 聚合物 数据挖掘 性能预测 反向设计 

分 类 号：T-19[一般工业技术]