Energy Storage Materials ( IF 18.9 ) Pub Date : 2019-12-17 , DOI: 10.1016/j.ensm.2019.12.010 Zhao Ding , Zhiqian Chen , Tianyi Ma , Chang-Tien Lu , Wenhui Ma , Leon Shaw
The prediction of hydrogen release ability is indispensable to evaluating hydrogen storage performance of -based mixtures before experimentation. To achieve this goal, ensemble machine learning is employed to automatically infer the relationship between factors (i.e., sample preparation, mixing conditions and operational variables) and target ( release amount), providing exceptional insight into hydrogen release ability. Specifically, the importance ranking of major variables for the hydrogen release of has been proposed for the first time based on the constructed uni-component catalysts database. We train our developed EoE model on 2,071 uni-component catalysts data and attempt to predict the hydrogen release amounts of doping with the unseen bi-component catalysts. The appealing results demonstrate the effectiveness and robustness of EoE. The procedure established in this study presents a novel approach for accelerating the research and development of hydrogen storage materials over various catalysts.
中文翻译:
预测氢的释放能力 集成机器学习的基于混合的混合物
氢释放能力的预测对于评估氢的储氢性能必不可少。 实验之前使用基于混合物的混合物。为了实现这一目标,采用集成机器学习来自动推断因素(即样品制备,混合条件和操作变量)与目标之间的关系(释放量),提供对氢气释放能力的出色洞察力。具体来说,氢释放的主要变量的重要性等级首次基于构建的单组分催化剂数据库提出了建议。我们在2,071个单组分催化剂数据上训练了我们开发的EoE模型,并试图预测氢的氢释放量用看不见的双组分催化剂掺杂。诱人的结果证明了EoE的有效性和鲁棒性。在这项研究中建立的程序提出了一种新颖的方法,以加速在各种催化剂上储氢材料的研究和开发。