基于马群优化的燃料电池极化曲线拟合算法

Fitting algorithm of polarization curve for fuel cell based on horse herd optimization

极化曲线可体现燃料电池的基本特性。在工程上,极化曲线一般用半经验公式来表述,其中的特征参数通过实验数据进行拟合得到。以提高拟合效果为目的,提出一种结合群优化算法的燃料电池极化曲线拟合方法。通过马群优化算法找出最小二乘算法各参数的拟合状态量,使最小二乘的拟合起始点、上下限设置更为合理,并在无开路电压的数据集上提高了9%的开路电压拟合精度。该算法在实际中具有较强的工程应用价值。

The basic characteristics of fuel cell could be characterized by polarization curve.In engineering,the polarization curve was usually characterized by a semi-empirical formula,the characteristic parameters in the semi-empirical formula were obtained by fitting the experimental data.Aiming at improving the fitting effect,a method of fuel cell polarization curve fitting combined with swarm optimization algorithm was proposed.The fitting state of each parameter of the original least squares algorithm was found out through the horse herd optimization algorithm,which made the setting of the fitting starting point,upper and lower limits of the least squares more reasonable,improved the open-circuit voltage fitting accuracy by 9%in the data set without open-circuit voltage.The algorithm had strong engineering application value in practice.