基于扩展卡尔曼滤波的动力电池SOC估计方法
SOC estimation of power battery based on extended Kalman filter
唐中信 1叶今禄 1易健彬 1邵帅 1陆一 1姚彤禹 1谭先琳1
作者信息
- 1. 广西大学机械工程学院,广西 南宁 530004
- 折叠
摘要
针对卡尔曼滤波方法估计锂离子电池荷电状态(state of charge,SOC)误差较大的问题,基于扩展卡尔曼滤波(extended Kalman filter,EKF)算法,建立锂离子电池二阶电阻电容(resistor capacitance,RC)电路模型,使用泰勒级数对非线性函数进行线性展开,采用MATLAB仿真估算动态应力测试、城市循环、混合脉冲功率、脉冲工况4 种不同工况对应的动力电池SOC曲线,并与安时积分法估算的SOC曲线进行对比分析.仿真结果表明:采用EKF算法估算锂离子电池SOC时,动态应力测试出现最大误差为 0.015,4 种工况均方差均在 0.003 2 以内;误差分布更稳定,曲线更可靠,采用EKF算法估算锂离子电池SOC的有效性与精确性更高.
Abstract
In order to solve the problem of large errors in state of charge(SOC)estimation by Kalman filter for lithium-ion battery,a model of a lithium-ion battery second-order resistance capacitance(RC)circuit is built based on extended Kalman filter(EKF)algorithm,which uses Taylor series to expand a nonlinear function linearly.The SOC curves of dynamic stress test,city cycle,mixed pulse power and pulse conditions are estimated by MATLAB simulation,the results are compared with the SOC curve estimated by ampere integration method.The results show that the maximum error of dynamic stress measurement is 0.015,the mean square deviation of the four lithium-ion battery is less than 0.003 2,the error distribution is more stable and the curve is more reliable,the EKF algorithm is more effective and accurate in estimating SOC of lithium-ion battery.
关键词
EKF/动力电池/SOC/仿真模拟Key words
EKF/power battery/SOC/simulation引用本文复制引用
基金项目
广西大学大学生创新创业训练计划项目(202110593212)
出版年
2024
NETL
NSTL
万方数据