Available capacity estimation of lithium-ion batteriesbased on the optimized support vector regression algorithm

doi:10.19799/j.cnki.2095-4239.2023.0387

Abstract

Abstract:

The current data-driven available capacity estimation algorithms for lithium-ion batteries encounter various challenges, including an inaccurate aging feature extraction, a low tracking accuracy of the available capacity decline trend, and a long time required for model parameter optimization. This study addresses these issues by proposing an optimized support vector machine regression algorithm that accurately estimates the available capacity of lithium-ion batteries. First, through an analysis of the aging data of a lithium-ion battery, the peak value of the battery capacity increment curve and the corresponding voltage of this peak value are extracted as the characteristic factors for the battery's aging state. The rationality of these characteristic factors is then analyzed using the Pearson correlation coefficient. The sparrow optimization algorithm is used to optimize the kernel function parameters of the support vector machine (SVM) regression algorithm, and the available battery capacity is accurately estimated based on the optimized SVM regression model. Finally, the advanced nature of the sparrow optimization algorithm in parameter optimization is verifiedby comparing different kernel parameter optimization algorithms. The model accuracy is verified by comparing it with the traditional SVM, Gaussian process regression, long short-term memory network, and other algorithms for estimating the available capacity. In conclusion, the proposed optimized support vector regression model effectively tracks the decline trajectory of lithium-ion batteries and accurately estimatestheir available capacity. It obtains better estimation results on different batteries, with the maximum estimation error of available capacity being less than 2%.

Key words: lithium-ion battery, availablecapacity estimation, support vector regression, sparrow search algorithm

CLC Number:

TM 911

Zheng CHEN, Yang CHEN, Jiangwei SHEN, Xuelei XIA, Shiquan SHEN, Renxin XIAO. Available capacity estimation of lithium-ion batteriesbased on the optimized support vector regression algorithm[J]. Energy Storage Science and Technology, 2023, 12(10): 3203-3213.

Figures/Tables 16

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Table 1

Fig. 6

Fig. 7

Fig. 8

Table 2

Fig. 9

Table 3

Fig. 10

Table 4

Fig. 11

Table 5

References 19

1	ZHANG C, YANG F, KE X Y, et al. Predictive modeling of energy consumption and greenhouse gas emissions from autonomous electric vehicle operations[J]. Applied Energy, 2019, 254: 113597.
2	陈峥, 李金元, 舒星, 等. 基于混合神经网络的锂离子电池健康状态估算[J]. 昆明理工大学学报(自然科学版), 2022, 47(3): 87-95.
	CHEN Z, LI J Y, SHU X, et al. State of health estimation for lithium-ion battery based on hybrid neural network[J]. Journal of Kunming University of Science and Technology (Natural Science), 2022, 47(3): 87-95.
3	林名强, 吴登高, 郑耿峰, 等. 基于表面温度和增量容量的锂电池健康状态估计[J]. 汽车工程, 2021, 43(9): 1285-1290, 1284.
	LIN M Q, WU D G, ZHENG G F, et al. Estimation method of state of health of lithium battery based on surface temperature and incremental capacity[J]. Automotive Engineering, 2021, 43(9): 1285-1290, 1284.
4	SONG Y C, LIU D T, LIAO H T, et al. A hybrid statistical data-driven method for on-line joint state estimation of lithium-ion batteries[J]. Applied Energy, 2020, 261: 114408.
5	谷平维, 段彬, 康永哲, 等. 基于随机充电数据的锂离子电池容量在线估计[J/OL]. 机械工程学报: 1-11[2023-04-28]. http://kns.cnki.net/kcms/detail/11.2187.TH.20230110.1451.021.html.
6	SHU X, LI G A, ZHANG Y J, et al. Online diagnosis of state of health for lithium-ion batteries based on short-term charging profiles[J]. Journal of Power Sources, 2020, 471: 228478.
7	GUO P Y, CHENG Z, YANG L. A data-driven remaining capacity estimation approach for lithium-ion batteries based on charging health feature extraction[J]. Journal of Power Sources, 2019, 412: 442-450.
8	LI X Y, YUAN C G, LI X H, et al. State of health estimation for Li-Ion battery using incremental capacity analysis and Gaussian process regression[J]. Energy, 2020, 190: 116467.
9	CHEN Z, SUN M M, SHU X, et al. Online state of health estimation for lithium-ion batteries based on support vector machine[J]. Applied Sciences, 2018, 8(6): 925.
10	LI X Y, WANG Z P, ZHANG L. Co-estimation of capacity and state-of-charge for lithium-ion batteries in electric vehicles[J]. Energy, 2019, 174: 33-44.
11	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/OL]. 高电压技术: 1-14[2023-04-28]. https://doi.org/10.13336/j.1003-6520.hve.20221843.
12	CHEN Z, XUE Q, WU Y T, et al. Capacity prediction and validation of lithium-ion batteries based on long short-term memory recurrent neural network[J]. IEEE Access, 2020, 8: 172783-172798.
13	COUTO L D, SCHORSCH J, JOB N, et al. State of health estimation for lithium ion batteries based on an equivalent-hydraulic model: An iron phosphate application[J]. Journal of Energy Storage, 2019, 21: 259-271.
14	申江卫, 马文赛, 肖仁鑫, 等. 基于优化高斯过程回归算法的锂离子电池可用容量估算[J]. 中国公路学报, 2022, 35(8): 31-43.
	SHEN J W, MA W S, XIAO R X, et al. Available capacity estimation of lithium-ion batteries based on optimized Gaussian process regression[J]. China Journal of Highway and Transport, 2022, 35(8): 31-43.
15	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.
	SHU X, LIU Y G, SHEN J W, 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.
16	韦荣阳, 毛阗, 高晗, 等. 基于TWP-SVR的锂离子电池健康状态估计[J]. 储能科学与技术, 2022, 11(8): 2585-2599.
	WEI R Y, MAO T, GAO H, et al. Estimation of health state of lithium-ion battery based on TWP-SVR[J]. Energy Storage Science and Technology, 2022, 11(8): 2585-2599.
17	张萍, 陆霞, 孟庆鹤. 基于多策略麻雀搜索算法的微电网容量优化配置[J]. 电气技术, 2023, 24(1): 1-9.
	ZHANG P, LU X, MENG Q H. Optimal allocation of micro-grid capacity based on multi-strategy sparrow search algorithm[J]. Electrical Engineering, 2023, 24(1): 1-9.
18	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.
19	李尧太. 基于AICKF的锂离子动力电池全寿命周期SOC估算研究[D]. 镇江: 江苏大学, 2021.
	LI Y T. Research on life cycle SOC estimation of lithium-ion power battery based on AICKF[D].Zhenjiang: Jiangsu University, 2021.

健康因子	电池编号
健康因子	电池1	电池2	电池3	电池4	电池5
F1	0.9746	0.9802	0.9697	0.9741	0.9754
F2	0.9593	0.9401	0.8924	0.9673	0.9493

评价标准	R²	MAE/%	AAE/%	RMSE/%	训练时间/s
50%	0.8969	4.28	0.64	1.11	1.2131
60%	0.9977	1.02	0.11	0.22	1.7697
70%	0.9982	0.99	0.08	0.19	3.1239

评价指标	不同优化算法
评价指标	SSA-SVR	GA-SVR	PSO-SVR
R²	0.9977	0.9667	0.9626
MAE/%	1.02	3.98	3.25
AAE/%	0.11	0.38	0.39
RMSE/%	0.22	0.83	0.88
优化时间/s	1.77	3.13	4.47

模型	评价标准	误差结果/%	优点	缺点
SSA-SVR	MAE	1.02	模型运行速度快，超参数寻优效果好	处理大样本数据时，精度略有下降
	AAE	0.11
	RMSE	0.22
SVM	MAE	8.48	小样本估算效果好	在处理大样本数据时效果差，超参数寻优效果差
	AAE	1.07
	RMSE	2.24
LSTM	MAE	1.96	擅长处理时序数据	需要大量数据，训练结果过于依赖样本质量
	AAE	0.28
	RMSE	0.56
GPR	MAE	6.75	适合高维度样本数据，具备不确定性表达能力	计算成本高，超参数寻优效果差
	AAE	1.25
	RMSE	2.41

评价指标	电池编号
评价指标	电池2	电池3	电池4	电池5
R²	0.9881	0.9821	0.9551	0.9672
MAE/%	1.77	1.95	1.65	1.92
AAE/%	0.30	0.47	0.95	0.49
RMSE/%	0.46	0.52	0.99	0.64