基于随机抽样参数辨识的动力锂电池组仿真模型构建

为解决多电芯电池组建模过程中电池参数辨识大量耗时的问题,根据电池组中各单体电池参数呈正态分布的特点,提出一种基于统计学方法和最小电动势的复合电池组建模方法.该方法首先随机抽取电池组中部分单体作为样本,利用脉冲充放电实验和最小二乘辨识方法对样本电池进行参数辨识;然后以样本电池参数的平均值和方差代表电池组的参数特征进行电池组模型构建;最后基于Simulink平台进行计算机仿真实验,并对比实际充放电实验数据.结果表明,两者的端电压误差在0.05 V以内,相对误差1.5%,验证了所提方法的有效性.

Li-ion batteries simulation modeling based on parameter identification on random sampling approach

In order to better estimate the state-of-charge lithium-ion batteries, it is necessary to establish an accurate battery pack model. Considering great time consumption in the process of identifying battery parameters for battery pack modeling, a creative method of battery pack modeling combining statistical method with minimum electromotive force is proposed in this paper. First, the parameters of concerned cells, sampling randomly from battery pack, are identified by practical experiments of pulse charging and discharging. The average and variance of the parameters of samples are obtained to represent the characteristics of target battery pack. Then, the simulation model of battery pack is established on the basis of Matlab/Simulink platform. Compared with discharge voltage results of the simulation and actual charge-discharge experiments on the whole battery pack , the difference is less than 0.05V, and the relative error is only 1.5%, which reveals the effectiveness of method suggested.