含能材料2025，Vol.33Issue(9)：981-992,12.DOI:10.11943/CJEM2025098
含能材料机器学习研究的数据优化策略
Data Optimization Strategies for Machine Learning of Energetic Materials
摘要
Abstract
As an emerging data-driven technology,machine learning provide a promising pathway for the intelligent research and development of energetic materials.However,data scarcity and heterogeneity have become the core bottleneck that restricts modeling accuracy and practical application.Focusing on the acquisition path and the existing of energetic material data,this re-view evaluates the mainstream data optimization strategies from two perspectives:quantity expansion and quality improvement.For data quantity expansion,recent advances in SMILES enumeration,generative adversarial networks,and transfer learning are introduced for enhancing model generalization ability.For data quality improvement,the roles of outlier detection,standardized preprocessing,and feature engineering in improving model robustness and interpretability are discussed.It is shown that effec-tive data optimization can not only alleviate data limitations but also significantly enhance prediction stability and structural ex-trapolation capabilities under small-sample and structurally complex conditions.Finally,future directions are proposed,includ-ing the development of high-throughput experimental platforms,unification of data standards,and establishment of intelligent closed-loop systems.It is expected to provide a feasible roadmap and methodological reference for advancing the data ecosystem and intelligent design of energetic materials.关键词
含能材料/机器学习/数据优化/模型泛化能力/模型鲁棒性Key words
energetic materials/machine learning/data optimization/model generalization/model robustness分类
军事科技引用本文复制引用
刘辰昊,张蕾,庞思平..含能材料机器学习研究的数据优化策略[J].含能材料,2025,33(9):981-992,12.基金项目
先进材料—国家科技重大专项(2024ZD0607000) （2024ZD0607000）
国家自然科学基金优秀青年科学基金项目(12222204) Advanced Materials-National Science and Technology Major Project(No.2024ZD0607000) （12222204）
National Natural Science Foundation of China(No.12222204) （No.12222204）
