采用非线性最小二乘法的超级电容等效电路模型参数辨识

A Parameter Identification Method for an Equivalent Circuit Model of Supercapacitor Using Nonlinear Least Squares

针对电动汽车混合储能系统中超级电容在传统模型参数辨识中误差较大、不利于估计其工作状态和进行系统能量管理的问题,提出了一种采用非线性最小二乘法的等效电路模型参数辨识方法。为了精确表征超级电容内部电荷再分配过程及其端电压特性,减小模型误差,该方法首先在超级电容三分支等效电路模型的基础上,分析超级电容充、放电过程中内部的电荷再分配过程;然后,利用粒子群算法分段辨识得到模型参数初值;最后,在所得模型参数初值的基础上采用基于信赖域的非线性最小二乘法对模型参数进行优化。采用超级电容进行多组恒流充、放电实验,结果表明,在不同充、放电条件下,模型计算的端电压与对应实验测试端电压的平均绝对百分误差均在0.5%以内,由此验证了基于非线性最小二乘法参数辨识的有效性和准确性。

A parameter identification method for supercapacitor equivalent circuit model using nonlinear least squares is designed to solve the problem that the model error of supercapacitors in hybrid energy storage system of electric vehicle is large and affects the state estimation of supercapacitors and system energy management.In order to actually characterize the process of charge redistribution inside the supercapacitor and its terminal voltage performance and to reduce the model error,a nonlinear parameter identification method is proposed.Firstly,based on a three branch equivalent circuit model,the initial parameters of the model are obtained by analyzing the internal charge redistribution process of the supercapacitor in the charge/discharge process.Then,initial values of model parameters are obtained by using a particle swarm optimization method accordingly.Finally,the nonlinear least-squares method with a trust region technique is employed to optimize the model parameters.The effectiveness and accuracy of the proposed parameter identification method based on the nonlinear least-squares method are verified by multiple constant current charge/discharge experiments for the supercapacitor.Results show that the average percentage errors between the terminal voltage calculated by the simulation model and the corresponding measured terminal voltage are within 0.5%under different charge/discharge conditions.