Scholarlink

<p>Proton exchange membrane electrolytic cells (PEMECs) have been recognized as a highly efficient and promising technology for sustainable hydrogen production. In this study, a comprehensive three-dimensional, two-phase, non-isothermal PEMEC model is developed to elucidate the intricate transport phenomena. The distributions of hydrogen and oxygen molar concentrations, liquid water volume fraction, and current density are investigated to reveal the transport characteristic and reaction mechanisms. In addition, a multi-objective optimization framework is implemented. A representative dataset is generated using latin hypercube sampling (LHS), and a back-propagation neural network optimized by the sparrow search algorithm (SSA-BP) is employed as a surrogate model. The non-dominated sorting genetic algorithm II (NSGA-II) is then used to obtain the Pareto front and then the optimal solution is selected by using the technique for order preference by similarity to an ideal solution (TOPSIS). The optimal operating condition parameters are 2.11&#xa0;V, 347.13&#xa0;K and 0.106&#xa0;m/s. This study provides a feasible strategy for the optimization of PEMECs operating parameters and offers valuable guidance for further improving cell performance while reducing trial-and-error costs.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Performance prediction and multi-objective optimization of PEMEC based on neural network and genetic algorithm

  • 期刊论文

  • 发表日期：

    2026年8月4日

查看全文
  • Dingding Xiao,
  • Shian Li,
  • Ruiyang Zhang,
  • Xinping Liu,
  • Aolong Liu,
  • Qiuwan Shen

摘要

Proton exchange membrane electrolytic cells (PEMECs) have been recognized as a highly efficient and promising technology for sustainable hydrogen production. In this study, a comprehensive three-dimensional, two-phase, non-isothermal PEMEC model is developed to elucidate the intricate transport phenomena. The distributions of hydrogen and oxygen molar concentrations, liquid water volume fraction, and current density are investigated to reveal the transport characteristic and reaction mechanisms. In addition, a multi-objective optimization framework is implemented. A representative dataset is generated using latin hypercube sampling (LHS), and a back-propagation neural network optimized by the sparrow search algorithm (SSA-BP) is employed as a surrogate model. The non-dominated sorting genetic algorithm II (NSGA-II) is then used to obtain the Pareto front and then the optimal solution is selected by using the technique for order preference by similarity to an ideal solution (TOPSIS). The optimal operating condition parameters are 2.11 V, 347.13 K and 0.106 m/s. This study provides a feasible strategy for the optimization of PEMECs operating parameters and offers valuable guidance for further improving cell performance while reducing trial-and-error costs.