中文摘要：

将具有全局搜索能力的遗传算法应用于质子交换膜燃料电池（PEMFC）扩散电极的性能优化,通过对PEMFC单体建立二维稳态数值计算模型,在ISIGHT-FD软件平台上利用径向基函数（RBF）神经网络拟和模型,在相应的设计空间内生成RBF拟和曲面,调用多岛遗传算法（MIGA）对RBF拟和进行遗传搜索,得到了阴极扩散层厚度、孔隙率和渗透率的最优值,通过优化前后的氧气浓度和输出性能比较,表明这些参数可改善气体扩散层的传质性能.

英文摘要：

The global searching Multi-Island Genetic Algorithm （MIGA） is used to optimize the performance of porous electrodes in a single proton exchange membrane fuel cell.A two-dimensional steady-state electrochemical mathematic model of a single PEMFC was established and implemented in Engineous’ ISIGHIT-FD.A global metamodel using radial basis functions （RBF） is built first in the domain bound.Using the metamodel,MIGA is then applied to determine the optimum value of the cathode gas diffusion layer thickness,porosity and permeability.By comparing the optimized oxygen mole fraction and polarization curve with the reference case,the cell performance was improved through the gas diffusion layer optimization.