基于iCEEMDAN和迁移学习的锂离子电池SOH估计*

doi:10.11985/2022.04.002

电气工程学报 ›› 2022, Vol. 17 ›› Issue (4): 2-10.doi: 10.11985/2022.04.002

• 特邀专栏：电化学储能系统安全管理与运维 • 上一篇下一篇

扫码分享

基于iCEEMDAN和迁移学习的锂离子电池SOH估计^*

杨淞元¹(), 田勇¹(), 田劲东¹^,²()

1.深圳大学物理与光电工程学院深圳 518060
2.人工智能与数字经济广东省实验室(深圳) 深圳 518132

收稿日期:2022-08-25 修回日期:2022-10-24 出版日期:2022-12-25 发布日期:2023-02-03
通讯作者: 田勇，男，1985年生，副教授，硕士研究生导师。主要研究方向为新能源汽车动力电池系统建模与状态估计，无线电能传输。E-mail：ytian@szu.edu.cn
作者简介:杨淞元，女，1998年生，硕士研究生。主要研究方向为基于数据驱动的锂离子电池健康状态预测。E-mail：2110456052@email.szu.edu.cn
田劲东，男，1973年生，教授，博士研究生导师。主要研究方向为人工智能技术及其应用，光学测量技术与仪器。E-mail：jindt@szu.edu.cn
基金资助:
*人工智能与数字经济广东省实验室(深圳)开放课题资助项目(GML-KF-22-19)

State of Health Estimation of Lithium-ion Batteries Based on iCEEMDAN and Transfer Learning

YANG Songyuan¹(), TIAN Yong¹(), TIAN Jindong¹^,²()

1. College of Physics and Optoelectronic Engineering, Shenzhen University, Shenzhen 518060
2. Guangdong Laboratory of Artificial Intelligence and Digital Economy(SZ), Shenzhen 518132

Received:2022-08-25 Revised:2022-10-24 Online:2022-12-25 Published:2023-02-03
Contact: TIAN Yong, E-mail：ytian@szu.edu.cn

摘要/Abstract

摘要：

目前数据驱动的锂离子电池健康状态(State of health，SOH)估计方法已成为研究热点，但实车应用中产生的小样本数据问题会导致数据驱动模型精度低、泛化能力差等问题，由此提出一种基于特征模态分解及迁移学习的SOH估计方法。首先，从电池小样本数据片段中提取健康特征，通过改进的自适应噪声完备集合经验模态分解(Improved complete ensemble empirical mode decomposition with adaptive noise，iCEEMDAN)分离出本征模态分量(Intrinsic mode function，IMF)与残余分量(Res)两类包含不同特征信息的分量；然后将分解优化后的特征信息分别通过LSTM网络和BP网络进行针对性训练，构建特征信息与电池SOH的关联模型；最后将模型迁移至其他数据集估计电池的SOH。基于NASA公开电池数据集的试验结果表明，所提方法具有高准确度及泛化能力，估计的平均绝对误差(MAE)和方均根误差(RMSE)分别为2.34%和3.05%，迁移后的MAE和RMSE分别为1.13%和1.68%。

关键词: 健康状态, iCEEMDAN, 数据驱动, 迁移学习, 小样本数据

Abstract:

The data-driven method for state of health(SOH) of lithium-ion batteries is currently a research hotspot. For electric vehicle applications, however, it has to face the challenge of small sample data, which leads to low accuracy and poor generalization. A SOH estimation method based on feature mode decomposition and transfer learning is proposed. Firstly, health features are extracted from a small section of the battery data set, and then they are divided into the intrinsic mode function(IMF) part and the residual signal(RES) part by using the improved complete ensemble empirical mode decomposition with adaptive noise(iCEEMDAN). Secondly, the IMF and RES parts are trained through a long short-term memory network and a back-propagation network, respectively, achieving a combined base model between the health features and the SOH. Finally, the base model is transferred to other data sets for SOH estimation. Validation results based on the NASA battery data set show that the proposed method performs high accuracy and generalization ability. The mean absolute error(MAE) and root mean square error(RMSE) are about 2.34% and 3.05%, respectively. With transfer learning, the MAE and RMSE are reduced to 1.13% and 1.68%, respectively.

Key words: Key words：State of health, iCEEMDAN, data-driven, transfer learning, small sample data

中图分类号:

TM912

杨淞元, 田勇, 田劲东. 基于iCEEMDAN和迁移学习的锂离子电池SOH估计^*[J]. 电气工程学报, 2022, 17(4): 2-10.

YANG Songyuan, TIAN Yong, TIAN Jindong. State of Health Estimation of Lithium-ion Batteries Based on iCEEMDAN and Transfer Learning[J]. Journal of Electrical Engineering, 2022, 17(4): 2-10.

图/表 12

图1

表1

表2

图2

图3

图4

图5

表3

图6

表4

图7

表5

参考文献 20

[1]	骆凡, 黄海宏, 王海欣. 基于电化学阻抗谱的退役动力电池荷电状态和健康状态快速预测[J]. 仪器仪表学报, 2021, 42(9):172-180.
	LUO Fan, HUANG Haihong, WANG Haixin. Rapid prediction of the state of charge and state of health of decommissioned power batteries based on electrochemical impedance spectroscopy[J]. Chinese Journal of Scientific Instrument, 2021, 42(9):172-180.
[2]	LI J, ADEWUYI K, LOTFI N, et al. A single particle model with chemical/mechanical degradation physics for lithium-ion battery state of health (SOH) estimation[J]. Applied Energy, 2018, 212:1178-1190. doi: 10.1016/j.apenergy.2018.01.011
[3]	匡柯, 孙跃东, 任东生, 等. 车用锂离子电池电化学-热耦合高效建模方法[J]. 机械工程学报, 2021, 57(14):10-22. doi: 10.3901/JME.2021.14.010
	KUANG Ke, SUN Yuedong, REN Dongsheng, et al. Efficient approach for electrochemical-thermal coupled modeling of large-format lithium-ion power battery[J]. Journal of Mechanical Engineering, 2021, 57(14):10-22. doi: 10.3901/JME.2021.14.010
[4]	贾俊, 胡晓松, 邓忠伟, 等. 数据驱动的锂离子电池健康状态综合评分及异常电池筛选[J]. 机械工程学报, 2021, 57(14):141-149,159. doi: 10.3901/JME.2021.14.141
	JIA Jun, HU Xiaosong, DENG Zhongwei, et al. Data-driven comprehensive evaluation of lithium-ion battery state of health and abnormal battery screening[J]. Journal of Mechanical Engineering, 2021, 57(14):141-149,159. doi: 10.3901/JME.2021.14.141
[5]	LI Xiaoyu, YUAN Changgui, LI Xiaohui, et al. State of health estimation for Li-ion battery using incremental capacity analysis and Gaussian process regression[J]. Energy, 2020, 190:116467. doi: 10.1016/j.energy.2019.116467
[6]	吴铁洲, 刘思哲, 张晓星, 等. 基于FA-BP神经网络的锂离子电池SOH估算[J]. 电池, 2021, 51(1):21-25.
	WU Tiezhou, LIU Sizhe, ZHANG Xiaoxing, et al. SOH estimation of Li-ion battery based on FA-BP neural network[J]. Battery Bimonthly, 2021, 51(1):21-25.
[7]	WU Yitao, XUE Qiaoshen, JIANG Wei, et al. State of health estimation for lithium-ion batteries based on healthy features and long short-term memory[J]. IEEE Access, 2020:28533-28547.
[8]	WANG Jiwei, DENG Zhongwei, YU Tao, et al. State of health estimation based on modified Gaussian process regression for lithium-ion batteries[J]. Journal of Energy Storage, 2022, 51:104512. doi: 10.1016/j.est.2022.104512
[9]	MA Yan, YAO Meihao, LIU Hongcheng, et al. State of health estimation and remaining useful life prediction for lithium-ion batteries by improved particle swarm optimization-back propagation neural network[J]. Journal of Energy Storage, 2022,52,104750.
[10]	GOH H H, LAN Zhentao, ZHANG Dongdong, et al. Estimation of the state of health (SOH) of batteries using discrete curvature feature extraction[J]. Journal of Energy Storage, 2022, 50:104646. doi: 10.1016/j.est.2022.104646
[11]	MA Yan, SHAN Ce, GAO Jinwu, et al. A novel method for state of health estimation of lithium-ion batteries based on improved LSTM and health indicators extraction[J]. Energy, 2022, 251:123973. doi: 10.1016/j.energy.2022.123973
[12]	WEN Jianping, CHEN Xing, LI Xianghe, et al. SOH prediction of lithium battery based on IC curve feature and BP neural network[J]. Energy, 2022, 261:125234. doi: 10.1016/j.energy.2022.125234
[13]	YANG Duo, ZHANG Xu, PAN Rui, et al. A novel gaussian process regression model for state-of-health estimation of lithium-ion battery using charging curve[J]. Journal of Power Sources, 2018, 384:387-395. doi: 10.1016/j.jpowsour.2018.03.015
[14]	TAN Yandan, ZHAO Guangcai. Transfer learning with long short-term memory network for state-of-health prediction of lithium-ion batteries[J]. IEEE Transactions on Industrial Electronics, 2020, 67(10):8723-8731. doi: 10.1109/TIE.2019.2946551
[15]	National Aeronautics and Space Administration Prognostics Center of Excellence. PCoE datasets[EB/OL]. [2022-10-25]. https://ti.arc.nasa.gov/tech/dash/groups/pcoe/prognostic-data-repository/.
[16]	FAN Lingfeng, WANG Ping, CHENG Ze. A remaining capacity estimation approach of lithium-ion batteries based on partial charging curve and health feature fusion[J]. Journal of Energy Storage, 2021, 43:103115. doi: 10.1016/j.est.2021.103115
[17]	COLOMINAS M A, SCHLOTTHAUER G, TORRES M E. Improved complete ensemble EMD:A suitable tool for biomedical signal processing[J]. Biomedical Signal Processing and Control, 2014, 14:19-29. doi: 10.1016/j.bspc.2014.06.009
[18]	LIANG Yanhui, LIN Yu, LU Qin. Forecasting gold price using a novel hybrid model with ICEEMDAN and LSTM-CNN-CBAM[J]. Expert Systems with Applications, 2022, 206:117847. doi: 10.1016/j.eswa.2022.117847
[19]	HE Y, KIM F T. Universities power energy management:A novel hybrid model based on iCEEMDAN and Bayesian optimized LSTM[J]. Energy Reports, 2021, 7:6473-6488. doi: 10.1016/j.egyr.2021.09.115
[20]	ZHANG Yongzhi, XIONG Rui, HE Hongwen, et al. Long short-term memory recurrent neural network for remaining useful life prediction of lithium-ion batteries[J]. IEEE Transactions on Vehicular Technology, 2018, 67(7):5695-5705. doi: 10.1109/TVT.2018.2805189

特征编号	特征含义
F1	片段充电电压区间容量
F2	片段充电所用时间
F3	单位时间内电压升
F4	ICA曲线峰值
F5	ICA曲线峰值对应电压

特征编号	电池数据集
特征编号	B5	B6	B7
F1	0.992 8	0.993 0	0.986 5
F2	0.992 6	0.993 0	0.986 1
F3	0.963 0	0.928 0	0.965 5
F4	0.991 0	0.991 8	0.985 9
F5	0.896 5	0.965 0	0.739 7

方法	RMSE	MAPE	MAE	R²
对照组1	0.044 8	18.808 0	0.039 1	0.973 0
对照组2	0.054 1	13.849 6	0.044 2	0.960 6
对照组3	0.086 3	40.249 1	0.074 8	0.899 8
对照组4	0.071 9	18.010 8	0.058 0	0.930 3
本文方法	0.030 5	9.807 7	0.023 4	0.987 4

方法	RMSE	MAPE	MAE	R²
EMD	0.057 6	18.857 5	0.050 0	0.955 4
CEEMDAN	0.059 2	15.120 3	0.042 3	0.952 8
iCEEMDAN	0.030 5	9.807 7	0.023 4	0.987 4

基于iCEEMDAN和迁移学习的锂离子电池SOH估计^*

State of Health Estimation of Lithium-ion Batteries Based on iCEEMDAN and Transfer Learning

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 20

相关文章 15

编辑推荐

Metrics

本文评价

[1]	邵海东, 肖一鸣, 颜深. 仿真数据驱动的改进无监督域适应轴承故障诊断[J]. 机械工程学报, 2023, 59(3): 76-85.
[2]	王震坡, 王秋诗, 刘鹏, 张照生. 大数据驱动的动力电池健康状态估计方法综述[J]. 机械工程学报, 2023, 59(2): 151-168.
[3]	王文东, 李杰, 张俊博, 秦雷, 袁小庆. 基于迁移学习的手部离散动作识别方法研究[J]. 机械工程学报, 2022, 58(7): 12-19.
[4]	徐彬翔, 郑林锋, 黄乙恒, 肖志能, 王新月. 基于改进最小二乘支持向量机的锂离子电池健康状态快速估计方法^*[J]. 电气工程学报, 2022, 17(4): 11-19.
[5]	方德宇, 楚潇, 刘涛, 李俊夫. 基于数据-模型驱动的锂离子电池健康状态估计^*[J]. 电气工程学报, 2022, 17(4): 20-31.
[6]	刘勇, 刘大鹏, 穆勇, 李振成, 王顺. 基于GWO优化ICEEMDAN分解的混合储能系统功率分配策略^*[J]. 电气工程学报, 2022, 17(4): 257-267.
[7]	宰红斌, 刘建国, 张文岗, 封士永, 祖国强. 基于深度迁移学习的配电线路绝缘子状态监测方法^*[J]. 电气工程学报, 2022, 17(4): 275-281.
[8]	徐茂舒, 沈旖, 王晟, 张娥, 李浩秒, 周敏, 王玮, 王康丽, 蒋凯. 先进感知技术在电池状态估计中的应用与启示^*[J]. 电气工程学报, 2022, 17(3): 40-57.
[9]	党月懋, 张雪纯, 徐楚奕, 江全元. 基于温度变化率曲线的锂离子电池健康状态评估算法^*[J]. 电气工程学报, 2022, 17(3): 58-65.
[10]	车云弘, 邓忠伟, 李佳承, 谢翌, 胡晓松. 基于数据驱动的电池系统泛化SOH估计方法[J]. 机械工程学报, 2022, 58(24): 253-263.
[11]	管寅昕, 杨吉祥, 丁汉. 考虑控制电信号中间量的机床运动控制误差数据驱动建模方法[J]. 机械工程学报, 2022, 58(23): 178-187.
[12]	王闰南, 谢帆, 张波. 基于神经网络方法的峰值电流控制Boost变换器数据驱动建模^*[J]. 电气工程学报, 2022, 17(2): 142-150.
[13]	鲍劲松, 张荣, 李婕, 陆玉前, 彭涛. 面向人-机-环境共融的数字孪生协同技术[J]. 机械工程学报, 2022, 58(18): 103-115.
[14]	易茜, 刘益君, 卓俊康, 李聪波, 易树平. BRBP-MOSOA融合数据驱动的热处理工艺低碳优化方法[J]. 机械工程学报, 2022, 58(16): 370-383.
[15]	邓柯楠, 高栋, 马守东, 路勇. 基于迁移学习的铣削机器人定位误差补偿方法[J]. 机械工程学报, 2022, 58(14): 170-180.