摘要
本文提出一种基于数据驱动法(data driven method,DDM)-等效电路模型(equivalent circuit model,ECM)融合的锂离子电池多时间尺度状态联合估计方法。首先提取内阻作为健康特征(health factor,HF),利用最小二乘支持向量机(least squares support vector machine,LSSVM)建立电池老化模型实现健康状态(state of health,SOH)估计;根据阻容参数辨识值和容量估计值建立电池状态空间方程,结合无迹卡尔曼滤波算法(unscented Kalman filter,UKF)进行荷电状态(state of charge,SOC)估计;用高斯过程回归(Gaussian process regression,GPR)对HF随循环次数的变化进行映射,预测HF的变化趋势,并结合LSSVM模型实现长期剩余使用使命(remaining useful life,RUL)预测。实验结果表明,所提方法具有较高精度和鲁棒性。
This paper proposes a multi-time scale joint state estimation method for lithium-ion batteries based on the fusion of data driven method(DDM)and equivalent circuit model(ECM).Firstly,the internal resistance is extracted as the health factor(HF),and the least squares support vector machine(LSSVM)is used to establish the battery aging model to estimate state of health(SOH).Subsequently,the battery state space equation is established according to the RC parameter identification values and the capacity estimation values,and then the Unscented Kalman Filter(UKF)algorithm is used to estimate state of charge(SOC).Then,Gaussian process regression(GPR)is used to map the change of HF with the number of cycles to predict the trend of HF change,and longterm remaining useful life(RUL)prediction is realized combined with the LSSVM model.The experimental results show that the proposed method has high accuracy and robustness.
作者
王萍
彭香园
程泽
张吉昂
Wang Ping;Peng Xiangyuan;Cheng Ze;Zhang Ji’ang(School of Electrical and Information Engineering,Tianjin University,Tianjin 300072)
出处
《汽车工程》
EI
CSCD
北大核心
2022年第3期362-371,378,共11页
Automotive Engineering
基金
国家自然科学基金(61873180)资助。
关键词
锂离子电池
多时间尺度
联合估计
融合方法
lithium-ion battery
multi-time scale
joint estimation
fusion method
作者简介
通信作者:程泽,教授,E-mail:chengze@tju.edu.cn。
