基于模型融合与自适应无迹卡尔曼滤波算法的锂离子电池SOC估计

摘要

In order to enhance the SOC estimation accuracy of power battery in electric vehicles,the lithi-um-ion battery model and its parameter identification algorithm,adaptive unscented Kalman filtering(AUKF) algo-rithm and the SOC estimation algorithm based on battery model fusion are studied in this paper. A battery circuit model with clear physical meanings is established,a AUKF-based battery SOC estimation scheme is devised by u-sing model parameter identification algorithm based on genetic algorithm,and a battery model fusion method is pro-posed based on Bayesian information criteria to fulfill battery SOC estimation based on model fusion. The results of simulation show the high accuracy of SOC estimation method proposed.%为提高电动汽车动力电池SOC的估计精度,本文中对锂离子电池模型与参数辨识算法、自适应无迹卡尔曼滤波(AUKF)算法和基于电池模型融合的SOC估计算法进行研究.建立了具有明确物理意义的电池电路模型,采用基于遗传算法(GA)的模型参数辨识算法,设计了基于AUKF的电池SOC估计方法,并基于贝叶斯信息准则,提出了电池模型融合方法,实现了基于模型融合与AUKF的电池SOC估计.仿真结果验证了该方法具有较高的精度.