基于核函数和超参数优化的退役锂电池健康状态估计

doi:10.19799/j.cnki.2095-4239.2023.0918

摘要/Abstract

摘要：

退役锂电池的健康状态(SOH)估计对于电池再利用和环境可持续性至关重要，考虑到电池退役前使用条件的不确定性，为进一步实现数据驱动方法对退役锂电池SOH的精确估计，本研究提出一种改进高斯过程回归(GPR)模型的SOH估计方法。首先，收集退役锂电池的循环充放电数据，在考虑温度影响的同时，使用容量增量分析(ICA)和电化学阻抗谱(EIS)等方法，获取统计健康特征来表征退役锂电池的老化特性，并使用Pearson相关系数对所选统计特征进行相关性分析，筛选出与SOH相关性高的健康特征，消除特征冗余性。然后，基于单一核函数学习老化特征能力有限和传统超参数寻优方法效率不足的特点，将线性核函数和对角平方指数核函数结合，以更好地适应电池SOH估计任务中的多样性，同时，使用鲸鱼算法(WOA)对估计模型的超参数进行优化，以确保最佳拟合效果，建立改进的GPR估计模型以提高估计的精确性。最后，采用NASA电池数据集中具有不同初始健康状况的四个不同电池，来验证所提出方法的有效性，结果表明，本文所提方法可以提供准确的SOH估计，其中平均绝对误差均小于1.75%且均方根误差均小于2.42%。

关键词: 退役锂电池, 健康状态, 鲸鱼算法, 核函数, 高斯过程回归

Abstract:

Given the uncertainty surrounding preretirement battery conditions, this study aims to advance the data-driven approach for accurately estimating retired lithium batteries' state of health (SOH). To achieve this, an enhanced SOH estimation method based on the Gaussian process regression (GPR) model is proposed. Initially, cyclic charge and discharge data from retired lithium batteries are gathered, and statistical health characteristics are derived to depict the aging properties. Methods such as capacity increment analysis (ICA) and electrochemical impedance spectroscopy (EIS) are employed to consider temperature effects. The Pearson correlation coefficient was used to assess the correlation between selected statistical features and health characteristics, identifying those highly correlated with SOH to eliminate the feature redundancy. Subsequently, acknowledging the limitations of individual kernel functions and conventional hyperparameter optimization techniques, a hybrid approach combining linear and diagonal square exponential kernel functions is introduced to better accommodate the diverse nature of battery SOH estimation tasks. The whale optimization algorithm (WOA) is then applied to optimize the hyperparameters of the estimation model, ensuring optimal fitting. This leads to the establishment of an improved GPR estimation model to enhance estimation accuracy. Finally, the effectiveness of the proposed method was validated using four different cells with varied initial health conditions from the NASA battery dataset. Results demonstrate the method's capability to provide accurate SOH estimation, with a mean absolute error below 1.75% and a root mean square error below 2.42%.

Key words: retired lithium-ion batteries, state of health, whale optimization algorithm, kernel functions, Gaussian process regression

中图分类号:

TM 912

李臣, 张会林, 张建平. 基于核函数和超参数优化的退役锂电池健康状态估计[J]. 储能科学与技术, 2024, 13(6): 2010-2021.

Chen LI, Huilin ZHANG, Jianping ZHANG. Estimated state of health for retired lithium batteries using kernel function and hyperparameter optimization[J]. Energy Storage Science and Technology, 2024, 13(6): 2010-2021.

图/表 16

表1

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

表2

图11

表3

四种不同核函数的SOH估计的统计误差"

序号	核函数	估计误差/%
		NO.5		NO.6		NO.7		NO.18
		MAE	RMSE	MAE	RMSE	MAE	RMSE	MAE	RMSE
1	$k L I N i s o (x, x')$	1.6	1.9	0.51	0.68	1.67	2.04	0.52	0.64
2	$k S E i s o (x, x')$	7.32	10.24	3.69	4.75	10.95	15.09	0.93	1.32
3	$k S E i s o_S E i s o (x, x')$	7.32	10.24	3.69	4.75	10.95	15.09	0.93	1.32
4	$k L I N i s o_S E i s o (x, x')$	2.82	3.79	0.33	0.48	1.53	1.96	0.3	0.42

表3

图12

图13

参考文献 29

1	李建林, 李雅欣, 陈光, 等. 退役动力电池健康状态特征提取及评估方法综述[J]. 中国电机工程学报, 2022, 42(4): 1332-1347.
	LI J L, LI Y X, CHEN G, et al. Research on feature extraction and SOH evaluation methods for retired power battery[J]. Proceedings of the CSEE, 2022, 42(4): 1332-1347.
2	CHENG M, ZHANG X, RAN A H, et al. Optimal dispatch approach for second-life batteries considering degradation with online SoH estimation[J]. Renewable and Sustainable Energy Reviews, 2023; 173: 113053.
3	CHEN J M, ZHANG W, GONG B G, et al. Optimal policy for the recycling of electric vehicle retired power batteries[J]. Technological Forecasting and Social Change, 2022, 183: 121930.
4	WANG J W, DENG Z W, TAO Y, et al. State of health estimation based on modified Gaussian process regression for lithium-ion batteries[J]. Journal of Energy Storage, 2022, 51: 104512.
5	CHOU J H, WANG F K, LO S C. Predicting future capacity of lithium-ion batteries using transfer learning method[J]. Journal of Energy Storage, 2023, 71: 108120.
6	WU J, LIU Z L, ZHANG Y, et al. Data-driven state of health estimation for lithium-ion battery based on voltage variation curves[J]. Journal of Energy Storage, 2023, 73: 109191.
7	LIU J Z, LIU X T. An improved method of state of health prediction for lithium batteries considering different temperature[J]. Journal of Energy Storage, 2023, 63: 107028.
8	ZHAO J, LI X B, YU D W, et al. Lithium-ion battery state of health estimation using meta-heuristic optimization and Gaussian process regression[J]. Journal of Energy Storage, 2023, 58: 106319.
9	XIONG X, WANG Y J, LI K Q, et al. State of health estimation for lithium-ion batteries using Gaussian process regression-based data reconstruction method during random charging process[J]. Journal of Energy Storage, 2023, 72: 108390.
10	LI Q L, LI D Z, ZHAO K, et al. State of health estimation of lithium-ion battery based on improved ant lion optimization and support vector regression[J]. Journal of Energy Storage, 2022, 50: 104215.
11	ZHANG Z P, MIN H T, GUO H G, et al. State of health estimation method for lithium-ion batteries using incremental capacity and long short-term memory network[J]. Journal of Energy Storage, 2023, 64: 107063.
12	LIU S Z, CHEN Z Q, YUAN L H, et al. State of health estimation of lithium-ion batteries based on multi-feature extraction and temporal convolutional network[J]. Journal of Energy Storage, 2024, 75: 109658.
13	FENG H L, LI N J. A multi-feature fusion model based on differential thermal capacity for prediction of the health status of lithium-ion batteries[J]. Journal of Energy Storage, 2023, 72: 108419.
14	DAI H D, WANG J X, HUANG Y Y, et al. Lightweight state-of-health estimation of lithium-ion batteries based on statistical feature optimization[J]. Renewable Energy, 2023, 119907.
15	FU S Y, TAO S Y, FAN H T, et al. Data-driven capacity estimation for lithium-ion batteries with feature matching based transfer learning method[J]. Applied Energy, 2024, 353: 121991.
16	BUCHICCHIO E, DE ANGELIS A, SANTONI F, et al. Battery SOC estimation from EIS data based on machine learning and equivalent circuit model[J]. Energy, 2023, 283: 128461.
17	ZHANG W C, LI T T, WU W X, et al. Data-driven state of health estimation in retired battery based on low and medium-frequency electrochemical impedance spectroscopy[J]. Measurement, 2023, 211: 112597.
18	ZHOU Z K, DUAN B, KANG Y Z, et al. Practical state of health estimation for LiFePO₄ batteries based on Gaussian mixture regression and incremental capacity analysis[J]. IEEE Transactions on Industrial Electronics, 2023, 70(3): 2576-2585.
19	ZUO H Y, LIANG J W, ZHANG B, et al. Intelligent estimation on state of health of lithium-ion power batteries based on failure feature extraction[J]. Energy, 2023, 282: 128794.
20	GUO Y F, YU X Y, WANG Y S, et al. Health prognostics of lithium-ion batteries based on universal voltage range features mining and adaptive multi-Gaussian process regression with Harris Hawks Optimization algorithm[J]. Reliability Engineering & System Safety, 2023, 109913.
21	王琛, 闵永军. 基于容量增量曲线与GWO-GPR的锂离子电池SOH估计[J]. 储能科学与技术, 2023, 12(11): 3508-3518.
	WANG C, MIN Y J. SOH estimation of lithium-ion batteries based on capacity increment curve and GWO-GPR[J]. Energy Storage Science and Technology, 2023, 12(11): 3508-3518.
22	陈锐, 丁凯, 祖连兴, 等. 基于AED-CEEMD-Transformer的锂离子电池健康状态估计[J]. 储能科学与技术, 2023, 12(10): 3242-3253.
	CHEN R, DING K, ZU L X, et al. Prediction of state of health of lithium-ion batteries based on the AED-CEEMD-Transformer network[J]. Energy Storage Science and Technology, 2023, 12(10): 3242-3253.
23	LU Z F, FEI Z C, WANG B F, et al. A feature fusion-based convolutional neural network for battery state-of-health estimation with mining of partial voltage curve[J]. Energy, 2024, 288: 129690.
24	ZHOU Y, DONG G Z, TAN Q Q, et al. State of health estimation for lithium-ion batteries using geometric impedance spectrum features and recurrent Gaussian process regression[J]. Energy, 2023, 262: 125514.
25	LIU S Z, NIE Y W, TANG A H, et al. Online health prognosis for lithium-ion batteries under dynamic discharge conditions over wide temperature range[J]. eTransportation, 2023, 18: 100296.
26	BAI J Q, HUANG J Y, LUO K, et al. A feature reuse based multi-model fusion method for state of health estimation of lithium-ion batteries[J]. Journal of Energy Storage, 2023, 70: 107965.
27	ZHOU Y F, WANG S L, XIE Y X, et al. Remaining useful life prediction and state of health diagnosis for lithium-ion batteries based on improved grey wolf optimization algorithm-deep extreme learning machine algorithm[J]. Energy, 2023, 285: 128761.
28	LI Y, WANG S L, CHEN L, et al. Multiple layer kernel extreme learning machine modeling and eugenics genetic sparrow search algorithm for the state of health estimation of lithium-ion batteries[J]. Energy, 2023, 282: 128776.
29	DONG H, MAO L, QU K Q, et al. State of health estimation and remaining useful life estimation for Li-ion batteries based on a hybrid kernel function relevance vector machine[J]. International Journal of Electrochemical Science, 2022, 17(11): 221135.

实验参数	B0005	B0006	B0007	B0018
额定容量/Ah	2	2	2	2
额定电压/V	3.7	3.7	3.7	3.7
电压上/下限/V	4.2/2.7	4.2/2.5	4.2/2.2	4.2/2.5
充/放电电流/A	1.5/2	1.5/2	1.5/2	1.5/2
温度/℃	24	24	24	24

电池编号	训练样本数	选取区间 (循环次数)	测试样本数	选取区间 (循环次数)
B0005	50	75～124	37	125～161
B0006	46	63～108	31	109～139
B0007	40	86～125	43	126～168
B0018	52	45～96	36	97～132

[1]	冯娜娜, 杨明, 惠周利, 王瑞洁, 宁弘扬. 基于蚁狮优化高斯过程回归的锂电池剩余使用寿命预测[J]. 储能科学与技术, 2024, 13(5): 1643-1652.
[2]	申小雨, 尹丛勃. 基于卷积Fastformer的锂离子电池健康状态估计[J]. 储能科学与技术, 2024, 13(3): 990-999.
[3]	朱亚宁, 张振东, 盛雷, 陈龙, 朱泽华, 付林祥, 毕青. 21700锂离子电池在不同健康状态下的热失控实验研究[J]. 储能科学与技术, 2024, 13(3): 971-980.
[4]	陈欣, 李云伍, 梁新成, 李法霖, 张志冬. 基于模态分解的Transformer-GRU联合电池健康状态估计[J]. 储能科学与技术, 2023, 12(9): 2927-2936.
[5]	李放, 闵永军, 张涌. 基于大数据的动力锂电池可靠性关键技术研究综述[J]. 储能科学与技术, 2023, 12(6): 1981-1994.
[6]	李林泽, 张向文. 基于组合频率阻抗特征的锂离子电池健康状态估算[J]. 储能科学与技术, 2023, 12(5): 1705-1712.
[7]	董渊昌, 庞晓琼, 贾建芳, 史元浩, 温杰, 李笑, 张鑫. 基于SVD-SAE-GPR的锂离子电池RUL预测[J]. 储能科学与技术, 2023, 12(4): 1257-1267.
[8]	赵鑫浩, 许亮. 改进的萤火虫算法优化反向传播神经网络动力锂离子电池健康状态估计[J]. 储能科学与技术, 2023, 12(3): 934-940.
[9]	王瑞洁, 惠周利, 杨明. 基于间接健康指标的高斯过程回归对锂电池SOH预测[J]. 储能科学与技术, 2023, 12(2): 560-569.
[10]	刘润兴, 盖玉成, 杨品哲, 张巍, 刘芹, 丁泽俊, 莫熙喆. 基于交流阻抗谱的铅酸蓄电池健康状态检测[J]. 储能科学与技术, 2023, 12(11): 3499-3507.
[11]	王琛, 闵永军. 基于容量增量曲线与GWO-GPR的锂离子电池SOH估计[J]. 储能科学与技术, 2023, 12(11): 3508-3518.
[12]	张宣梁, 何霆, 朱文龙, 王屾, 曾建华, 徐泉, 牛迎春. 基于多循环特征的储能电池SOH估计模型[J]. 储能科学与技术, 2023, 12(11): 3488-3498.
[13]	耿陈, 孟锦豪, 彭乔, 刘天琪, 曾雪洋, 陈刚. 基于弛豫过程特征提取的锂离子电池健康状态估计[J]. 储能科学与技术, 2023, 12(11): 3479-3487.
[14]	毛百海, 覃吴, 肖显斌, 郑宗明. 基于LSTM&GRU-Attention多联合模型的锂离子电池SOH估计[J]. 储能科学与技术, 2023, 12(11): 3519-3527.
[15]	陈锐, 丁凯, 祖连兴, 许青松, 王宗标, 罗大思, 苏敬江, 胡圣, 毛冀龙. 基于AED-CEEMD-Transformer的锂离子电池健康状态估计[J]. 储能科学与技术, 2023, 12(10): 3242-3253.