充电模态下电动汽车动力电池模型辨识

摘要

The battery model and parameters identification are the base of the charge and discharge optimal control of electric vehicle traction batteries.And the parameters of the battery model are affected by the working condition of the traction battery.In order to model the traction battery and identify the model parameters, modeling algorithm of traction battery and parameters identification method were studied in this paper.A variable order equivalent circuit model was established, which is based on the electrode impedance spectrum theory.A parameter identification algorithm was proposed, which is based on the forgetting factor recursive extended least square (FFRELS).A selection algorithm for the optimal order of the battery model was built, which is based on the Bayesian information criterions (BIC).A battery open circuit voltage model that was used for calibration of the proposed parameter identification algorithm was created, which is based on the lattice gas model (LGM).In the end, the battery model parameters identification algorithm and the optimal order selection under the charging mode was achieved.Validation results show that the proposed modeling and parameters identification algorithm is efficient.%电池模型及参数辨识是电动汽车动力电池进行充、放电优化控制的基础,同时模型参数受充、放电工况的影响.为对充电模态下的电动汽车动力电池进行建模与参数辨识,对动力电池建模方法、模型参数辨识算法展开研究,建立基于电极阻抗谱理论的可变阶次电池等效电路模型,提出基于遗忘因子扩展递推最小二乘算法(FFRELS)的模型参数辨识算法,构建基于贝叶斯信息准则(BIC)的电池模型最优阶次选择算法,创建基于晶格气体模型(LGM)的电池开路电压模型,对电池模型参数辨识算法进行修正,实现了充电模态下的电池模型参数辨识与最优阶次选择.仿真结果证明了该方法的有效性.