期刊
ENERGY CONVERSION AND MANAGEMENT
卷 226, 期 -, 页码 -出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.enconman.2020.113513
关键词
PEM fuel cell; Porous media flow field; Data-driven surrogate model; Machine learning
资金
- National Key Research and Development Program of China [2017YFB0102703]
- China-UK International Cooperation and Exchange Project (Newton Advanced Fellowship)
- National Natural Science Foundation of China [51861130359]
- UK Royal Society [NAF\R1\180146]
- Natural Science Foundation for Outstanding Young Scholars of Tianjin [18JCJQJC46700]
- Shanghai Everpower Technologies Ltd
Flow field design plays a key role in proton exchange membrane (PEM) fuel cells due to its decisive influence on reactant gas transfer, liquid water discharge, electron conductance, and heat transfer. Porous media flow field (e. g. metal foam) shows promise to improve cell performance in the concentration polarization regime and to benefit uniform distributions of oxygen, liquid, current density, and temperature, etc. In this study, a data-driven surrogate model based on support vector machine (SVM) is applied to optimize porous media flow field geometry. The training data is obtained from a validated three-dimensional (3D), multi-phase non-isothermal model. For better characterization, the complex structure is simplified as a 3D structured mesh consisting of a set of fibers with each fiber perpendicular to each other at the joint. The results show that the data-driven surrogate model based on SVM predicts similar results as the 3D physical model. The genetic algorithm (GA) is further used for optimization, in which the data-driven surrogate model is selected as a fitness evaluation function. The optimal values obtained by the surrogate model are verified by the 3D physical model, indicating that the proposed data-driven surrogate model is effective in the design and optimization of porous media flow field.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据