电源世界
2018/08
|
15
Research and Design
研究与设计
结合电路等效和 UKF-EKF 的锂电池
SOC
估算
方法研究
Study on
SOC
Estimation Method of Lithium Battery with
Circuit Equivalent and UKF-EKF
刘小菡
1
,王顺利
2
,苏杰
1
,顾鹏程
1
,马程
1
1. 西南科技大学信息工程学院(四川,绵阳,621010)
2. 检测研究所(四川,绵阳,621010)
Liu Xiaohan
1
, Wang Shunli
2
*, Su Jie
1
, Gu Pengcheng
1
, Ma Cheng
1
1.Southwest University of Science and Technology
,
(Mianyang, 621010, China)
2.School of Informaon Engineering SiChuan
,
(Mianyang, 621010, China)
摘 要:动力锂电池具备能量高、绿色环保等优点,逐步在汽车、航天、船舶等领域得以应用,常以成组形式被使用,以满足高电压、
大容量的供电需求。但在生产与使用中单体不一致现象严重影响了电池的使用效率、寿命以及安全性,电池管理系统(Battery
Management System, BMS)应运而生。动力电池荷电状态 (State of Charge,
SOC
) 的准确估算,对电动汽车电池管理系统 (Battery
Management System, BMS) 有极其重要的意义。针对动力电池估算所存在的等效模型模拟电池充放电过程中真实低、常用算法精
度损失等问题,基于二阶戴维南等效电路模型,对比分析扩展卡尔曼滤波(Extended Kalman Filter,EKF)算法与无迹卡尔曼滤
波 (Unscented Kalman Filter,UKF) 算法优、缺点,提出 UKF-EKF 联合的
SOC
在线估算。以 18490 三元锂电池为检测对象,利
用递推最小二乘法进行电池模型的参数辨识,将估计的
SOC
结果和试验测量结果进行比较,通过实验与仿真验证该方法的精度。
实验结果表明,
SOC
估计精度在 3% 内,为锂电池等效电路模型的分析研究提供了一种可行方式。
关键词:18490 三元锂电池 二阶戴维南等效电路模型 UKF-EKF 递推最小二乘法
Abstract: Power lithium batteries have the advantages of high energy, environmental protection and other advantages, and are gradually applied
in the automotive, aerospace, marine and other fields. They are often used in groups to meet the high-voltage, large-capacity power supply needs.
However, the inconsistency of monomer in production and use has seriously affected the efficiency, lifespan and safety of the battery. The battery
management system (BMS) came into being. Accurate estimation of the state of charge (
SOC
) of the power battery is important for the Battery
Management System (BMS) of the electric vehicle. Aimed at the problems of the low authenticity of the equivalent model , and the low accuracy
of the common algorithm, this paper took the 18490 ternary lithium battery as the test object, studied second-order Thevenin equivalent circuit
model. Based on the second-order Thevenin equivalent circuit model, Comparing the advantages and disadvantages of Extended Kalman Filter
(EKF) algorithm and Unscented Kalman Filter (UKF) algorithm. a UKF-EKF combined
SOC
online estimation was proposed. The 18490 ternary
lithium battery was used as the detection target. The parameters of the battery model were identified by using recursive least squares method. The
estimated
SOC
results and experimental measurements were obtained. The results were compared and the accuracy of the method was verified by
experiments and simulations. Experimental results show that the accuracy of
SOC
estimation is within 3%, which provides a feasible method for
the analysis and study of the equivalent circuit model of lithium battery.
Keywords: 18490 Ternary lithium battery, Second-order Thevenin equivalent circuit model, UKF-EKF, Recursive least squares method
[中图分类号]TM912.9 [文献标识码] A 文章编号:1561-0349(2018)08-0015-04
1 引言
针对锂电池组
SOC
估算与等效模型构建问题,国内外研
究者提出了大量解决方案。Zhang 等人提出基于小波变换矩
阵(WTM)的去噪方法,可对非平稳电流和电压信号进行
有效地分析和去噪,其
SOC
估计误差被限制在 1%以内
[1]
;
Zhang 等人通过神经网络在静态
SOC
基础上精确估计动态电