• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2019, Vol. 55 ›› Issue (20): 52-59.doi: 10.3901/JME.2019.20.052

• 特邀专栏:动力电池系统关键技术 • 上一篇    

基于充电曲线特征的锂离子电池容量估计

戴海峰1,2, 姜波1,2, 魏学哲1, 张艳伟3   

  1. 1. 同济大学汽车学院 上海 201804;
    2. 同济大学智能型新能源汽车协同创新中心 上海 201804;
    3. 上汽大众汽车有限公司 上海 201805
  • 收稿日期:2019-03-04 修回日期:2019-06-25 发布日期:2020-01-07
  • 通讯作者: 戴海峰(通信作者),男,1981年出生,博士,教授,博士研究生导师。主要研究方向为汽车电子、动力电池成组及管理、车载充电机以及新能源汽车硬件在环测试系统。E-mail:tongjidai@tongji.edu.cn
  • 作者简介:姜波,男,1992年出生,博士研究生。主要研究方向为动力电池管理。E-mail:jiangbo15@tongji.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(51677136,U1764256)。

Capacity Estimation of Lithium-ion Batteries Based on Charging Curve Features

DAI Haifeng1,2, JIANG Bo1,2, WEI Xuezhe1, ZHANG Yanwei3   

  1. 1. School of Automotive Studies, Tongji University, Shanghai 201804;
    2. Collaborative Innovation Center for Intelligent New Energy Vehicles, Tongji University, Shanghai 201804;
    3. SAIC Volkswagen Automobile Co., Ltd., Shanghai 201805
  • Received:2019-03-04 Revised:2019-06-25 Published:2020-01-07

摘要: 准确的容量估计对锂离子电池管理具有重要意义。通过电池循环老化试验,归纳出老化过程中与电池容量衰减相关的充电曲线特征。通过计算充电曲线特征与衰减容量的相关系数进一步确定特征的电压区间。建立以径向基函数为核函数的相关向量机模型,以5个特征为输入量、电池容量为输出量进行数据训练,然后以筛选出的相关矢量进行在线容量估计。结果表明,该电池容量估计算法精度在2.2%以内,且算法收敛性较好。

关键词: 锂离子电池, 容量估计, 充电曲线特征, 相关系数, 相关向量机

Abstract: Accurate capacity estimation plays an important role in lithium-ion battery management. The charging curve features related to the battery capacity attenuation during the aging process are summarized through the battery cycle aging experiments. By calculating the correlation coefficient between curve features and attenuation capacity, the voltage range of curve features is further determined. A relevance vector machine with radial basis function as the kernel function is established. Five features are adopted as input and battery capacity as output for data training, and then the trained relevant sparse vector is used for online capacity estimation. The estimation results show that the accuracy of the data-driven capacity estimation algorithm is less than 2.2% and the convergence of the algorithm is rapid.

Key words: lithium-ion batteries, capacity estimation, charging curve feature, correlation coefficient, relevance vector machine

中图分类号: