基于GPR与NSGA-Ⅱ的爪式氢气循环泵多目标结构优化

车用发动机 ›› 2026, Vol. 0 ›› Issue (4) : 87-94.

李霞^1,2，胡海波¹，高琳¹，朱斌³，刘本学^1,2

作者信息 +

Multi-Objective Structural Optimization of Claw-Type Hydrogen Circulation Pump Based on GPR and NSGA-Ⅱ#br#

LI Xia^1,2,HU Haibo¹,GAO Lin¹,ZHU Bin³,LIU Benxue^1,2

Author information +

文章历史 +

摘要

爪式氢气循环泵作为质子交换膜燃料电池（PEMFC）的核心组件，其性能直接影响燃料电池的整体效能。为提高爪式氢气循环泵的供氢量和工作效率，对其结构开展了多目标优化设计。首先，基于高斯过程回归（GPR）建立爪式氢气循环泵平均流量与容积效率的预测模型。然后，将该模型与非支配排序遗传算法(NSGA-Ⅱ)结合，对平均流量与容积效率进行优化。最后，将Pareto前沿解集代入多准则决策（MCDM）分析方法，以寻找最优结构参数组合。结果表明：基于高斯过程回归建立的预测模型的决定系数大于0.95，均方根误差小于0.5，具有良好的预测精度和相关性，实际应用中可以极大地减少仿真与试验次数。使用NSGA-Ⅱ对结构参数优化后的爪式氢气循环泵的平均流量相对增长了12.04%，容积效率相对增长了2.68%。

Abstract

As a core component of the proton exchange membrane fuel cell (PEMFC), the performance optimization of claw-type hydrogen circulation pump is crucial for improving the overall efficiency of fuel cell. To enhance the hydrogen supply capacity and working efficiency of claw-type hydrogen circulation pump, multi-objective optimization design was carried out for its structure. A prediction model for the average flow rate and volumetric efficiency of claw-type hydrogen circulation pump was first established based on Gaussian process regression (GPR). Then the model was combined with the non-dominated sorting genetic algorithm-Ⅱ (NSGA-Ⅱ) to optimize average flow rate and volumetric efficiency, and the Pareto frontier solution set was substituted into the multi-criteria decision making (MCDM) analysis method to find the optimal combination of structural parameters. The results show that the prediction model based on GPR has a coefficient of determination greater than 0.95 and a root mean square error (RMSE) less than 0.5, which exhibits excellent prediction accuracy and correlation, and can greatly reduce the number of simulation and experiment in practical applications. Compared with the parameters before optimization, the optimal structural parameters obtained by NSGA-Ⅱ have increased the average flow rate of claw-type hydrogen circulation pump by 12.04% and the volumetric efficiency by 2.68% relatively.

导出引用

李霞, 胡海波, 高琳, 朱斌, 刘本学. 基于GPR与NSGA-Ⅱ的爪式氢气循环泵多目标结构优化[J]. 车用发动机. 2026, 0(4): 87-94

LI Xia, HU Haibo, GAO Lin, ZHU Bin, LIU Benxue. Multi-Objective Structural Optimization of Claw-Type Hydrogen Circulation Pump Based on GPR and NSGA-Ⅱ#br#[J]. Vehicle Engine. 2026, 0(4): 87-94