基于1stOpt的质子交换膜燃料电池模型优化  

Proton exchange membrane fuel cell model optimization based on 1stOpt

作　　者：王恺尔[1] 叶美盈[1] 许友生[1] 

机构地区：[1]浙江师范大学数理与信息工程学院,浙江金华321004

出　　处：《电源技术》2009年第8期698-700,共3页Chinese Journal of Power Sources

基　　金：国家自然科学基金(10572130)

摘　　要：质子交换膜燃料电池(PEMFC)的性能受温度,压力等众多外界因素以及本身结构的影响,具有较强的非线性,目前已建立的数学模型存在众多难以确定的参数。将1stOpt中的麦夸特法+通用全局优化法用于PEMFC的极化曲线模型,得到最优参数,避免了使用其他优化方法因为初始值选取不当对模型精确度造成的影响。通过对实验数据和加噪仿真数据的参数优化,结果表明这种方法建立的数学模型和实验数据之间能够达到很高的拟合精度,因而为PEMFC模型参数优化提出了一种高效实用的新方法。Proton exchange membrane fuel cell （PEMFC） is strongly nonlinear, since it is affected by temperature, pressure and some external factors as well as its structure. Furthermore, in PEMFC mathematical model there are many parameters difficult to be optimized by traditional technology as least square. The Levenberg-Marquardt ＋ Universal Global Optimization of lstOpt was applied to optimize a set of parameters in the PEMFC polarization curve model, avoiding the troubles and imprecise modeling result from inappropriate initial value guesses. Through simulation and experiment, the results show that lstOpt has an excellent modeling ability, which makes it an ideal, practical and efficient method in PEMFC model optimization.

关 键 词：质子交换膜燃料电池 建模 优化 

分 类 号：TM911[电气工程—电力电子与电力传动]