针对基于机器学习的锂离子电池电池健康状态（SOH）估计可解释性差、对数据的依赖性强等问题,提出一种具有可解释性的锂电池SOH估计方法。首先,研究分析随着电池老化,充电电压与首次充电电压之间的距离呈现出较好的趋势性,基于Shapelets概念提出能够捕捉电池退化趋势的距离度量,进一步通过相关性分析确定Shapelets候选集范围以提高特征提取效率,结合减法平均优化算法（SABO）的BP模型,进行Shapelets的选择。最后,设计基于Shapelets的SABO-BP模型,实现电池SOH的有效估计。在斯坦福大学与麻省理工学院（Stanford-MIT）提供的数据集上对所提方法进行验证,通过使用不同充电策略的电池进行测试,得到电池SOH估计的平均绝对误差（MAE）均维持在0.5%以内,最低可达0.19%;均方根误差（RMSE）也均保持在0.6%以内,最低可达0.26%;决定系数R2保持在0.98以上,最高可达0.995。实验结果表明,所提方法能在数据有限的情况下准确预测锂电池SOH,证实了所提算法的泛化性与实用价值。
Abstract
In response to the problems of poor interpretability and strong dependence on data for battery state of health（SOH） estimation of lithium-ion batteries based on machine learning, this paper proposes a SOH estimation method for lithium batteries with interpretability. First of all, this work analyzes that as batteries age, the distance between the charging voltage and the first charging voltage exhibits a good trend, and proposes a distance metric based on the concept of Shapelets that can capture the degradation trend of the battery, and further determines the range of the Shapelets candidate set through correlation analysis to improve the efficiency of the feature extraction. The selection of Shapelets is carried out in combination with the BP model optimized by the subtraction-average-based optimizer (SABO) algorithm. Finally, the Shapelets-based SABO-BP model is designed to realize the effective estimation of battery SOH. The proposed method is validated on the Stanford-MIT dataset, and batteries with different charging strategies are selected for test, the MAE of the battery SOH estimation are all maintained within 0.5%, and can reach as low as 0.19%; the RMSE are also all maintained within 0.6%, and can reach as low as 0.26%; and the R2 stays above 0.98 and reaches up to 0.995. The experimental results show that the proposed method is able to accurately predict the SOH of lithium batteries with limited data, which confirms the generalization and practical value of the proposed algorithm.
关键词
锂离子电池 /
状态估计 /
神经网络模型 /
Shapelets /
减法平均优化算法
Key words
lithium-ion batteries /
state estimation /
neural network models /
Shapelets /
subtraction-average-based optimizer
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 莫易敏, 余自豪, 叶鹏, 等. 基于迁移学习与GRU神经网络结合的锂电池SOH估计[J]. 太阳能学报, 2024, 45(3): 233-239.
MO Y M, YU Z H, YE P, et al.Lithium battery soh estimation method based on combination of transfer learning and GRU neural network[J]. Acta energiae solaris sinica, 2024, 45(3): 233-239.
[2] 谢文强. 遗传算法优化BP网络的锂电池剩余容量预测[J]. 仪表技术, 2019(1): 35-37, 40.
XIE W Q.A GA-based optimal algorithm for the BP network for forecasting remaining capacity of the lithium battery[J]. Instrumentation technology, 2019(1): 35-37, 40.
[3] XU T T, PENG Z, WU L F.A novel data-driven method for predicting the circulating capacity of lithium-ion battery under random variable current[J]. Energy, 2021, 218: 119530.
[4] 颜湘武, 邓浩然, 郭琪, 等. 基于自适应无迹卡尔曼滤波的动力电池健康状态检测及梯次利用研究[J]. 电工技术学报, 2019, 34(18): 3937-3948.
YAN X W, DENG H R, GUO Q, et al.Study on the state of health detection of power batteries based on adaptive unscented Kalman filters and the battery echelon utilization[J]. Transactions of China Electrotechnical Society, 2019, 34(18): 3937-3948.
[5] 寇发荣, 杨天祥, 罗希, 等. 基于特征重构与多时间尺度的锂电池SOH和RUL联合估计[J].太阳能学报,2025, 46(6): 68-78.
KOU F R,YANG T X, LUO X, et al.Joint estimation of SOH and RUL for lithium batteries based on feature reconstruction and multiple times scales[J]. Acta energiae solaris sinica, 2025, 46(6): 68-78.
[6] 何冰琛, 杨薛明, 王劲松, 等. 基于PCA-GPR的锂离子电池剩余使用寿命预测[J]. 太阳能学报, 2022, 43(5): 484-491.
HE B C, YANG X M, WANG J S, et al.Prediction of remaining useful life of lithium-ion batteries based on PCA-GPR[J]. Acta energiae solaris sinica, 2022, 43(5): 484-491.
[7] 陈晓宇, 耿萌萌, 王乾坤, 等. 基于电化学阻抗特征选择和高斯过程回归的锂离子电池健康状态估计方法[J]. 储能科学与技术, 2022, 11(9): 2995-3002.
CHEN X Y, GENG M M, WANG Q K, et al.Electrochemical impedance feature selection and Gaussian process regression based on the state-of-health estimation method for lithium-ion batteries[J]. Energy storage science and technology, 2022, 11(9): 2995-3002.
[8] 李谦, 姜帆, 韩乔妮, 等. 基于一阶ECM-IGPR的锂离子电池SOC及SOH联合估计框架[J]. 太阳能学报, 2024, 45(5): 240-250.
LI Q, JIANG F, HAN Q N, et al.Joint first order SOC and SOH estimation framework for li-ion battery based on ECM-IGPR[J]. Acta energiae solaris sinica, 2024, 45(5): 240-250.
[9] YU Q Q, NIE Y W, GUO S S, et al.Machine learning enables rapid state of health estimation of each cell within battery pack[J]. Applied energy, 2024, 375: 124165.
[10] 熊庆, 邸振国, 汲胜昌. 锂离子电池健康状态估计及寿命预测研究进展综述[J]. 高电压技术, 2024, 50(3):1182-1195.
XIONG Q, DI Z G,JI S C.Review on health state estimation and life prediction of lithium-ion batteries[J]. High voltage engineering, 2024,50(3): 1182-1195.
[11] 蔡艳平, 陈万, 苏延召, 等. 锂离子电池剩余寿命预测方法综述[J]. 电源技术, 2021, 45(5): 678-682.
CAI Y P, CHEN W, SU Y Z, et al.Review of remaining useful life prediction for lithium ion batteries[J]. Chinese journal of power sources, 2021, 45(5): 678-682.
[12] NAGULAPATI V M, LEE H, JUNG D, et al.A novel combined multi-battery dataset based approach for enhanced prediction accuracy of data driven prognostic models in capacity estimation of lithium ion batteries[J]. Energy and AI, 2021, 5: 100089.
[13] WANG L Z, JIANG S Y, MAO Y T, et al.Lithium-ion battery state of health estimation method based on variational quantum algorithm optimized stacking strategy[J]. Energy reports, 2024, 11: 2877-2891.
[14] 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.
ZHOU D, SONG X H, LU W B, 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.
[15] 王萍, 张吉昂, 程泽, 等. 基于充电电压片段的锂离子电池状态联合估计方法[J]. 湖南大学学报(自然科学版), 2021, 48(10): 187-200.
WANG P, ZHANG J A, CHENG Z, et al.A coupled state estimation method of lithium batteries based on partial charging voltage segment[J]. Journal of Hunan University (natural sciences), 2021, 48(10): 187-200.
[16] HE Y, BAI W Y, WANG L L, et al.SOH estimation for lithium-ion batteries: an improved GPR optimization method based on the developed feature extraction[J]. Journal of energy storage, 2024, 83: 110678.
[17] YE L X, KEOGH E.Time series shapelets: a new primitive for data mining[C]//Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Paris, France, 2009: 947-956.
[18] 燕晗. 基于shapelet的小样本滚动轴承故障诊断方法研究[D]. 无锡: 江南大学, 2022.
YAN H.Reaserch for few-shot fault diagnosis of rolling bearing based on shapelet[D]. Wuxi: Jiangnan University, 2022.
[19] YU H Y, XU C J, GENG G C, et al.Multi-time-scale shapelet-based feature extraction for non-intrusive load monitoring[J]. IEEE transactions on smart grid, 2023, 15(1): 1116-1128.
[20] YAN L J, LIU Y S, LIU Y.Application of discrete wavelet transform in shapelet-based classification[J]. Mathematical problems in engineering, 2020, 2020: 6523872.
[21] SEVERSON K A, ATTIA P M, JIN N, et al.Data-driven prediction of battery cycle life before capacity degradation[J]. Nature energy, 2019, 4(5): 383-391.
[22] YE M, WANG Q, YAN L S, et al.Enhanced robust capacity estimation of lithium-ion batteries with unlabeled dataset and semi-supervised machine learning[J]. Expert systems with applications, 2024, 238: 121892.
[23] ZHU J G, WANG Y X, HUANG Y, et al.Data-driven capacity estimation of commercial lithium-ion batteries from voltage relaxation[J]. Nature communications, 2022, 13: 2261.
[24] TROJOVSKÝ P, DEHGHANI M.Subtraction-average-based optimizer: a new swarm-inspired metaheuristic algorithm for solving optimization problems[J]. Biomimetics, 2023, 8(2): 149.
基金
国家自然科学基金（52307242）; 中央高校基本科研业务费专项资金（2023JC001）
PDF(2614 KB)
