基于ABC-SVM固体氧化物燃料电池电堆建模与仿真

Modeling and Simulation of Solid Oxide Fuel Cell Stack Based on ABC-SVM

为了更好地满足工程上对SOFC(solid oxide fuel cell)性能预测和控制方案设计要求,提出利用人工蜂群算法(ABC)优化支持向量机(SVM)来建立SOFC电堆模型。通过利用ABC算法优化SVM参数(核函数值宽度和惩罚因子),采用优化后的参数作为SVM的初始参数建立模型,与SVM、GA-SVM和PSO-SVM模型进行对比。实验结果表明:ABC-SVM模型平均平方误差小,说明该算法可以很好的预测在不同氢气流速下SOFC的电压/电流特性曲线。该模型对SOFC预测和控制方案设计有一定价值。

In order to better meet the engineering requirements for SOFC performance prediction and control scheme design,an artificial bee colony algorithm(ABC)optimization support vector machine(SVM)is proposed to establish the SOFC stack model.By using the ABC algorithm to optimize the SVM parameters(the kernel function with and the penalty coefficient),the optimized parameters are used as the initial parameters of the SVM.The model is compared with the SVM,GA-SVM and PSO-SVM models.The experimental results show that the ABC-SVM model s average squared error is small,which indicates that the algorithm can predict the voltage/current characteristic curve of SOFC under different hydrogen flow rates.The model has certain value for SOFC prediction and control scheme design.