基于数据驱动的电池系统泛化SOH估计方法

车云弘; 邓忠伟; 李佳承; 谢翌; 胡晓松

doi:10.3901/JME.2022.24.253

基于数据驱动的电池系统泛化SOH估计方法

doi: 10.3901/JME.2022.24.253

车云弘^{1, 2,},
邓忠伟^{1, 2},
李佳承^{1, 2},
谢翌^{1, 2,*, ,},
胡晓松^{1, 2}

1.
重庆大学机械与运载工程学院重庆 400044
2.
重庆大学机械传动国家重点实验室重庆 400044

基金项目:

国家重点研发计划 2022YFE0102700

国家自然科学基金 52111530194

重庆市英才计划 cstc2021ycjh-bgzxm0295

广东省重点领域研发计划 2020B0909030001

详细信息

作者简介:
车云弘，男，1996年出生。主要研究方向为电池状态估计及寿命预测。E-mail：yunhongche@126.com

通讯作者:
谢翌(通信作者)，男，1983年出生，博士，副教授，博士研究生导师。主要研究方向为电动汽车电池管理技术。E-mail：claudexie@cqu.edu.cn

中图分类号: TM912
计量
- 文章访问数: 23
- HTML全文浏览量: 38
- PDF下载量: 0
- 被引次数: 0
出版历程
- 收稿日期: 2022-01-25
- 修回日期: 2022-06-25
- 网络出版日期: 2024-03-07
- 刊出日期: 2022-12-20

Generalized Data-driven SOH Estimation Method for Battery Systems

CHE Yunhong^{1, 2
,},
DENG Zhongwei^{1, 2},
LI Jiacheng^{1, 2},
XIE Yi^{1, 2,*
, ,},
HU Xiaosong^{1, 2}

1.
College of Mechanical and Vehicle Engineering, Chongqing University, Chongqing 400044
2.
State Key Laboratory of Mechanical Transmissions, Chongqing University, Chongqing 400044

摘要

摘要: 准确可靠的电池健康状态估计是保证锂离子电池安全运行的关键，同时为失效预警提供参考。提出一种适用于电池单体和电池组的健康状态估计通用方法。首先，提出基于局部充放电数据的电池单体高效健康因子提取方法，保证健康因子和容量的高相关性和实现健康因子的在线可获取性。其次，提出考虑电池组容量衰减和不一致性的特征生成策略，利用主成分分析获取融合特征，利用双时间尺度滤波和电池组等效电路模型拓宽特征提取方法的应用范围。然后，基于高斯过程回归算法框架，考虑健康因子和容量衰减的整体关系和局部变化提出改进的高斯核函数提高估计精度和可靠性。最后，利用多个试验数据集验证算法在不同应用条件下的泛化能力。估计结果表明，对恒流放电工况的电池单体估计误差小于1.28%，在动态变温条件下电池单体估计误差小于1.82%；串联电池组的验证结果表明在各种应用场景下估计误差均小于1.43%。提高了电池系统健康状态估计的精度以及在广泛应用场景下的适应性。

准确可靠的电池健康状态估计是保证锂离子电池安全运行的关键，同时为失效预警提供参考。提出一种适用于电池单体和电池组的健康状态估计通用方法。首先，提出基于局部充放电数据的电池单体高效健康因子提取方法，保证健康因子和容量的高相关性和实现健康因子的在线可获取性。其次，提出考虑电池组容量衰减和不一致性的特征生成策略，利用主成分分析获取融合特征，利用双时间尺度滤波和电池组等效电路模型拓宽特征提取方法的应用范围。然后，基于高斯过程回归算法框架，考虑健康因子和容量衰减的整体关系和局部变化提出改进的高斯核函数提高估计精度和可靠性。最后，利用多个试验数据集验证算法在不同应用条件下的泛化能力。估计结果表明，对恒流放电工况的电池单体估计误差小于1.28%，在动态变温条件下电池单体估计误差小于1.82%；串联电池组的验证结果表明在各种应用场景下估计误差均小于1.43%。提高了电池系统健康状态估计的精度以及在广泛应用场景下的适应性。
- 锂离子电池 /
- 健康状态估计 /
- 健康因子 /
- 高斯过程回归 /
- 状态估计
Abstract: Accurate and reliable battery state of health estimation is the key to ensuring the safe operation of lithium-ion batteries, and provides a reference for failure warning. A general method to estimate the state of health of both battery cells and battery packs is proposed. Firstly, a method for extracting high-quality health indicators of battery cells based on partial charge or discharge data is proposed to ensure the high correlation between health indicators and battery capacity and the online availability of the health indicators. Secondly, a feature generation strategy that considers the capacity attenuation and inconsistency of the battery pack is proposed. The final fusion feature is extracted by using principal component analysis to reduce the dimensionality of the feature matrix. The dual time scale filtering and battery pack equivalent circuit model are combined to broaden the extraction under dynamic discharge conditions. Then, based on the framework of the Gaussian process regression, an improved Gaussian kernel function is proposed considering the overall relationship and local changes of the health indicators and capacity attenuation. Finally, multiple experimental data sets are used to verify the generalization ability of the proposed method under different application conditions. The estimation results show that the proposed method has an estimation error of less than 1.28% for battery cells under constant current discharge conditions, and an estimation error of less than 1.82% for battery cells under dynamic working conditions with changeable environmental temperatures. The verification results for series battery packs show that it can be used in various application scenarios with estimation errors all less than 1.43%. The accuracy of and adaptability in a wide range of application scenarios of battery state of health estimation for battery systems are improved.
- lithium-ion battery /
- state of health /
- health indicators /
- Gaussian process regression /
- state estimation

HTML全文

下载: 全尺寸图片幻灯片

图 1 动态老化工况电压响应

下载: 全尺寸图片幻灯片

图 2 公开数据集169个电池容量及HIs衰减曲线

下载: 全尺寸图片幻灯片

图 3 B1-B3电池估计结果及误差

下载: 全尺寸图片幻灯片

图 4 B4-B6电池估计结果及误差

下载: 全尺寸图片幻灯片

图 5 C1~C4电池估计结果及误差

下载: 全尺寸图片幻灯片

图 6 Pack#1~Pack#4电池估计结果及误差

下载: 全尺寸图片幻灯片

表 1 电池组规格参数及测试条件

明细	Pack#1	Pack#2	Pack#3	Pack#4
电极材料	NCM/石墨	LiFePO₄/石墨
额定容量	177 A·h	100 A·h	100 A·h	100 A·h
电压范围	2.8 V~4.2 V	2.6 V~3.6 V
充电模式	MCC	CC (0.5C)-CV (0.05C)
放电模式	CC (1C)	CC (0.5C)
电芯数量	4	16	15	15
容量范围/(A·h)	147.79~182.60	77.04~100.35	78.86~101.38	59.14~73.29
环境温度	25 ℃	35 ℃

下载: 导出CSV

表 2 公开数据集相关性分析统计结果

HI	均值	> 0.9比例	> 0.95比例	> 0.98比例
std_Q	0.968	0.982	0.899	0.361
std_dQ	0.972	0.982	0.888	0.491

下载: 导出CSV

表 3 试验数据集相关性分析结果

HI	C1	C2	C3	C4
std_Q	0.998	0.989	0.999	0.960
std_dQ	0.996	0.993	0.998	0.938

下载: 导出CSV

表 4 电池组HIs和容量的相关性分析结果

HI	pack#1	pack#2		Pack#3	Pack#4
HI	pack#1	normal	dynamic	Pack#3	Pack#4
PCA (std_Q)	0.998	0.997	0.984	0.991	0.995
PCA (std_△Q)	0.995	0.997	0.996	0.987	0.967

下载: 导出CSV

表 5 B1-B6电池使用不同方法的估计误差对比 %

电池编号	组合核函数		指数核函数		线性核函数
电池编号	MAE	RMSE	MAE	RMSE	MAE	RMSE
B1	0.555	0.676	0.636	0.731	0.565	0.682
B2	0.623	0.735	0.662	0.749	0.664	0.749
B3	0.577	0.856	0.589	0.874	0.577	0.857
B4	0.980	1.272	1.006	1.306	1.102	1.332
B5	0.486	0.714	0.492	0.835	0.513	0.809
B6	0.308	0.412	0.429	0.523	0.309	0.413

下载: 导出CSV

表 6 试验数据集SOH估计误差

电池编号	MAE (%)	RMSE (%)
C1	0.801	0.908
C2	1.040	1.140
C3	0.357	0.454
C4	1.427	1.811

下载: 导出CSV

表 7 电池组SOH不同方法的估计误差对比 %

电池编号	组合核函数		指数核函数		线性核函数
电池编号	MAE	RMSE	MAE	RMSE	MAE	RMSE
Pack#1	0.491	0.529	0.661	0.976	0.724	1.065
Pack#1(dynamic)	0.623	0.735	0.640	0.739	0.833	0.915
Pack#2	0.198	0.240	0.638	0.785	0.935	1.039
Pack#3	0.928	0.981	3.061	3.155	1.983	2.461
Pack#4	1.275	1.425	1.529	1.698	1.503	1.740

下载: 导出CSV

参考文献(26)

[1]	唐小林, 李珊珊, 王红, 等. 网联环境下基于分层式模型预测控制的车队能量控制策略研究[J]. 机械工程学报, 2020, 56(14): 119-128. doi: 10.3901/JME.2020.14.119 TANG Xiaolin, LI Shanshan, WANG Hong, et al. Research on energy control strategy based on hierarchical model predictive control in connected environment[J]. Journal of Mechanical Engineering, 2020, 56(14): 119-128. doi: 10.3901/JME.2020.14.119
[2]	杨瑞鑫, 熊瑞, 孙逢春. 锂离子动力电池外部短路测试平台开发与试验分析[J]. 电气工程学报, 2021, 16(1): 103-118. https://www.cnki.com.cn/Article/CJFDTOTAL-DQZH202101014.htm YANG Ruixin, XIONG Rui, SUN Fengchun. Experimental platform development and characteristics analysis of external short circuit in lithium-ion batteries[J]. Journal of Electrical Engineering, 2021, 16(1): 103-118. https://www.cnki.com.cn/Article/CJFDTOTAL-DQZH202101014.htm
[3]	CHE Yunhong, DENG Zhongwei, LI Penghua, et al. State of health prognostics for series battery packs: A universal deep learning method[J]. Energy, 2022, 238: 121857. doi: 10.1016/j.energy.2021.121857
[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]	HU Xiaosong, FENG Fei, LIU Kailong, et al. State estimation for advanced battery management: key challenges and future trends[J]. Renewable and Sustainable Energy Reviews, 2019, 114: 109334. doi: 10.1016/j.rser.2019.109334
[6]	周頔, 宋显华, 卢文斌, 等. 基于日常片段充电数据的锂电池健康状态实时评估方法研究[J]. 中国电机工程学报, 2019, 39(1): 105-111, 325. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGDC201901012.htm ZHOU Di, SONG Xianhua, LU Wenbin, et al. Real-time SOH estimation algorithm for lithium-ion batteries based on daily segment charging data[J]. Proceedings of the CSEE, 2019, 39(1): 105-111, 325. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGDC201901012.htm
[7]	周子游, 刘永刚, 杨阳, 等. 考虑混杂充电数据的锂离子电池容量估计[J]. 机械工程学报, 2021, 57(14): 1-9. doi: 10.3901/JME.2021.14.001 ZHOU Ziyou, LIU Yonggang, YANG Yang, et al. Capacity estimation of lithium-ion battery considering hybrid charging data[J] Journal of Mechanical Engineering, 2021, 57(14): 1-9. doi: 10.3901/JME.2021.14.001
[8]	HU Xiaosong, XU Le, LIN Xianke, et al. Battery lifetime prognostics[J]. Joule, 2020, 4(2): 310-46.
[9]	LYU Zhiqiang, GAO Renjing, CHEN Lin. Li-ion battery state of health estimation and remaining useful life prediction through a model-data-fusion method[J]. IEEE Transactions on Power Electronics, 2021, 36(6): 6228-6240.
[10]	HU Xiaosong, CHE Yunhong, LIN Xianke, et al. Battery health prediction using fusion-based feature selection and machine learning[J]. IEEE Transactions on Transportation Electrification, 2021, 7(2): 382-398.
[11]	CHE Yunhong, DENG Zhongwei, LIN Xianke, et al. Predictive battery health management with transfer learning and online model correction[J]. IEEE Transactions on Vehicular Technology, 2021, 70(2): 1269-1277.
[12]	GOU Bin, XU Yan, FENG Xue. State-of-health estimation and remaining-useful-life prediction for lithium-ion battery using a hybrid data-driven method[J]. IEEE Transactions on Vehicular Technology, 2020, 69(10): 10854-10867.
[13]	WENG Caihao, FENG Xuning, SUN Jing, et al. State-of-health monitoring of lithium-ion battery modules and packs via incremental capacity peak tracking[J]. Applied Energy, 2016, 180: 360-8.
[14]	舒星, 刘永刚, 申江卫, 等. 基于改进最小二乘支持向量机与Box-Cox变换的锂离子电池容量预测[J]. 机械工程学报, 2021, 57(14): 118-128. doi: 10.3901/JME.2021.14.118 SHU Xing, LIU Yonggang, SHEN Jiangwei, et al. Capacity prediction for lithium-ion batteries based on improved least squares support vector machine and Box-Cox transformation[J], Journal of Mechanical Engineering, 2021, 57(14): 118-128. doi: 10.3901/JME.2021.14.118
[15]	SUI Xin, HE Shan, VILSEN S B, et al. A review of non-probabilistic machine learning-based state of health estimation techniques for Lithium-ion battery[J]. Applied Energy, 2021, 300: 117346.
[16]	HU Xiaosong, CHE Yunhong, LIN Xianke, et al. Health prognosis for electric vehicle battery packs: A data-driven approach[J]. IEEE/ASME Transactions on Mechatronics, 2020, 25(6): 2622-32.
[17]	SEVERSON Kristen A., ATTIA Peter M., JIN Norman, et al. Data-driven prediction of battery cycle life before capacity degradation[J]. Nature Energy, 2019, 4(5): 383-391.
[18]	ATTIA Peter M, GROVER Aditya, JIN Norman, et al. Closed-loop optimization of fast-charging protocols for batteries with machine learning[J]. Nature, 2020, 578(7795): 397-402.
[19]	DENG Zhongwei, HU Xiaosong, LIN Xianke, et al. General discharge voltage information enabled health evaluation for lithium-ion batteries[J]. IEEE/ASME Transactions on Mechatronics, 2021, 26(3): 1295-1306.
[20]	CHE Yunhong, DENG Zhongwei, TANG Xiaolin, et al. Lifetime and aging degradation prognostics for lithium-ion battery packs based on a cell to pack method[J]. Chinese Journal of Mechanical Engineering, 2022, 35(4): 1-16.
[21]	CHE Yunhong, FOLEY Aoife, El-GINDY Moustafa, et al. Joint estimation of inconsistency and state of health for series battery packs[J]. Automotive Innovation, 2021(4): 103-116.
[22]	王萍, 范凌峰, 程泽. 基于健康特征参数的锂离子电池SOH和RUL联合估计方法[J]. 中国电机工程学报, 2022, 42(4): 1523-1534. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGDC202204024.htm WANG Ping, FAN Lingfeng, CHENG Ze. A joint state of health and remaining useful life estimation approach for lithium-ion batteries based on health factor parameter[J]. Proceedings of the CSEE, 2021, 42(4): 1523-1534. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGDC202204024.htm
[23]	高铭琨, 徐海亮, 吴明铂. 基于等效电路模型的动力电池SOC估计方法综述[J]. 电气工程学报, 2021, 16(1): 90-102. https://www.cnki.com.cn/Article/CJFDTOTAL-DQZH202101013.htm GAO Mingkun, XU Hailiang, WU Mingbo. Review of SOC estimation methods for power battery based on equivalent circuit model[J]. Journal of Electrical Engineering, 2021, 16(1): 90-102. https://www.cnki.com.cn/Article/CJFDTOTAL-DQZH202101013.htm
[24]	FENG Fei, HU Xiaosong, LIU Kailong, et al. A practical and comprehensive evaluation method for series-connected battery pack models[J]. IEEE Transactions on Transportation Electrification, 2020, 6(2): 391-416.
[25]	HU Lin, HU Xiaosong, CHE Yunhong, et al. Reliable state of charge estimation of battery packs using fuzzy adaptive federated filtering[J]. Applied Energy, 2020, 262: 114569.
[26]	胡晓松, 唐小林. 电动车辆锂离子动力电池建模方法综述[J]. 机械工程学报, 2017, 53(16): 20-31. doi: 10.3901/JME.2017.16.020 HU Xiaosong, TANG Xiaolin. Review of modeling techniques for lithium-ion traction batteries in electric vehicles[J] Journal of Mechanical Engineering, 2017, 53(16): 20-31. doi: 10.3901/JME.2017.16.020