Applying machine learning to balance performance and stability of high energy density materials.
Applying machine learning to balance performance and stability of high energy density materials.
复制标题
DOI：
10.1016/j.isci.2021.102240
复制
发表时间：
2021-03-19
期刊：
影响因子：
5.8
通讯作者：
Zhang L
中科院分区：
文献类型：
--
作者：
Huang X;Li C;Tan K;Wen Y;Guo F;Li M;Huang Y;Sun CQ;Gozin M;Zhang L
高能量密度材料（HEDMs）长期存在的性能-稳定性矛盾问题具有极其复杂和多参数的性质。在本文中，机器学习被用于处理153个HEDM的28个特征描述符和5个爆炸和稳定性特性，其中所有21，648个数据都是通过超级计算机上的高通量晶体级量子力学计算获得的。采用分层抽样和5折交叉验证方法分别对极端梯度Boosting回归树（XGBoost）、自适应Boosting、随机森林、多层感知器和核岭回归5种模型进行训练和评估。其中，XGBoost模型在预测HEDMs的爆速、爆压、爆炸热、分解温度和晶格能方面产生了最好的评分指标，并且XGBoost预测与大量文献中收集的1，383个实验数据最吻合。进行特征重要性分析以获得对性能-稳定性矛盾的因果关系的数据驱动洞察，并提供关键特征的最佳范围，以实现先进HEDM的更高效合理设计。对153个HEDM的21，648个物理化学数据进行晶体级量子力学计算HEDM的爆轰和稳定性的机器学习和实验验证以数据为驱动力洞察性能-稳定性矛盾的因果关系先进HEDM合理设计的关键特征的最佳范围计算材料科学;材料科学计算方法;能源材料;材料设计
The long-standing performance-stability contradiction issue of high energy density materials (HEDMs) is of extremely complex and multi-parameter nature. Herein, machine learning was employed to handle 28 feature descriptors and 5 properties of detonation and stability of 153 HEDMs, wherein all 21,648 data used were obtained through high-throughput crystal-level quantum mechanics calculations on supercomputers. Among five models, namely, extreme gradient boosting regression tree (XGBoost), adaptive boosting, random forest, multi-layer perceptron, and kernel ridge regression, were respectively trained and evaluated by stratified sampling and 5-fold cross-validation method. Among them, XGBoost model produced the best scoring metrics in predicting the detonation velocity, detonation pressure, heat of explosion, decomposition temperature, and lattice energy of HEDMs, and XGBoost predictions agreed best with the 1,383 experimental data collected from massive literatures. Feature importance analysis was conducted to obtain data-driven insight into the causality of the performance-stability contradiction and delivered the optimal range of key features for more efficient rational design of advanced HEDMs. Crystal-level quantum mechanics calculations of 21,648 physicochemical data of 153 HEDMs Machine learning of detonation and stability of HEDMs and experimental validation Data-driven insight into the causality of performance-stability contradiction Optimal range of key features for rational design of advanced HEDMs Computational Materials Science; Computational Method in Materials Science; Energy Materials; Materials Design
登录
查看更多内容
影响因子：
1.8
作者：
Crawford, Margaret-Jane;Evers, Juergen;Welch, Jan M.
通讯作者：
Welch, Jan M.
影响因子：
3.8
作者：
Averkiev, BB;Antipin, MY;Sheremetev, AB
通讯作者：
Sheremetev, AB
影响因子：
3.6
作者：
Altmann, KL;Chafin, AP;Gilardi, R
通讯作者：
Gilardi, R
影响因子：
3.6
作者：
ATKINS, RL;HOLLINS, RA;WILSON, WS
通讯作者：
WILSON, WS
影响因子：
3.7
作者：
Antonangeli, Daniele;Farber, Daniel L.;Klepeis, John E.
通讯作者：
Klepeis, John E.