基于卷积Fastformer的锂离子电池健康状态估计

doi:10.19799/j.cnki.2095-4239.2023.0735

摘要/Abstract

摘要：

锂离子电池的健康状态（state of health，SOH）是电池管理系统的重要功能，对于电池的可靠运行和使用寿命具有重要意义。为了进一步提高数据驱动方法对锂离子电池SOH估计的精度，提出一种卷积Fastformer模型的SOH估计方法。首先，提取锂离子电池多个充电阶段的每次循环电压曲线、电流曲线，每个阶段各个曲线转换为统计健康特征来表征锂离子电池老化特性，并使用Pearson相关系数对所选统计特征进行了相关性分析，筛选出与容量相关性高的健康特征，消除特征冗余性。随后，融合卷积神经网络和具有线性复杂度的Fastformer神经网络的特点，使用卷积神经网络强大的特征提取能力挖掘健康特征的局部信息，利用Fastformer的多头附加注意力机制可以更高效地在复杂的长序列中总结全文信息。然后，为减少模型训练时间，利用正交实验法对模型超参数进行优化。最后，采用公开数据集将所提方法与CNN、GRU、RNN模型进行对比，验证卷积Fastformer模型的准确性，结果表明，平均绝对误差、均方根误差最大仅为0.25%，0.29%，相对误差在0.8%以内，具有较高的估计精度和稳定性。

关键词: 锂离子电池, 健康状态估计, 正交实验, 卷积Fastformer

Abstract:

The state of health (SOH) of lithium-ion battery batteries is a crucial parameter for battery management systems, playing a substantial role in ensuring reliable operation and extending battery lifespan. This study aims to introduce an estimation method based on convolutional Fastformer model to improve the accuracy of data-driven SOH estimation for lithium-ion batteries. Initially, voltage and current curves from various charging stages of the lithium-ion battery are extracted for each cycle. These curves are then transformed into statistical health features, providing insights into the battery's aging characteristics. The Pearson correlation coefficient is used to analyze the relationship between selected statistical features and capacity, facilitating the identification of highly correlated health features and the elimination of redundant ones. Based on the strengths of convolutional neural networks and the linear complexity of Fastformer neural network, our approach combines the feature extraction capability of the former with local information mining of health features. The Fastformer's multihead attention mechanism efficiently summarizes contextual information within lengthy sequences. Model training time is reduced by optimizing hyperparameters using an orthogonal experimental method. Finally, a publicly available dataset is used for comparative evaluations, pitting our approach against other models such as CNN, GRU, and RNN. The results validate the accuracy of the convolutional Fastformer model, with maximum mean absolute error and root mean square error at only 0.25% and 0.29%, respectively, and a relative error within 0.08%. These findings demonstrate the high accuracy and stability achieved by the proposed method for SOH estimation.

Key words: lithium-ion battery, SOH estimation, orthogonal experiment, convolutional Fastformer

中图分类号:

TM 912

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.

Xiaoyu SHEN, Congbo YIN. SOH estimation of lithium-ion batteries using a convolutional Fastformer[J]. Energy Storage Science and Technology, 2024, 13(3): 990-999.

图/表 15

图1

表1

图2

图3

图4

图5

图6

图7

图8

图9

表2

表3

图10

表4

图11

参考文献 24

1	ZHANG Y, LI Y F. Prognostics and health management of Lithium-ion battery using deep learning methods: A review[J]. Renewable and Sustainable Energy Reviews, 2022, 161: 112282.
2	TIAN J Q, LIU X H, LI S Q, et al. Lithium-ion battery health estimation with real-world data for electric vehicles[J]. Energy, 2023, 270: 126855.
3	LIN M Q, WU D G, MENG J H, et al. Health prognosis for lithium-ion battery with multi-feature optimization[J]. Energy, 2023, 264: 126307.
4	SHAO J Y, LI J F, YUAN W Z, et al. A novel method of discharge capacity prediction based on simplified electrochemical model-aging mechanism for lithium-ion batteries[J]. Journal of Energy Storage, 2023, 61: 106788.
5	YANG L, CAI Y S, YANG Y X, et al. Supervisory long-term prediction of state of available power for lithium-ion batteries in electric vehicles[J]. Applied Energy, 2020, 257: 114006.
6	KIM S, PARK H J, CHOI J H, et al. A novel prognostics approach using shifting kernel particle filter of Li-ion batteries under state changes[J]. IEEE Transactions on Industrial Electronics, 2021, 68(4): 3485-3493.
7	ZHANG Y H, YANG Y, XIU X C, et al. A remaining useful life prediction method in the early stage of stochastic degradation process[J]. IEEE Transactions on Circuits and Systems II: Express Briefs, 2021, 68(6): 2027-2031.
8	DONG G Z, HAN W J, WANG Y J. Dynamic Bayesian network-based lithium-ion battery health prognosis for electric vehicles[J]. IEEE Transactions on Industrial Electronics, 2021, 68(11): 10949-10958.
9	LIU K L, SHANG Y L, OUYANG Q, et al. A data-driven approach with uncertainty quantification for predicting future capacities and remaining useful life of lithium-ion battery[J]. IEEE Transactions on Industrial Electronics, 2021, 68(4): 3170-3180.
10	GOH H H, LAN Z T, ZHANG D D, et al. Estimation of the state of health (SOH) of batteries using discrete curvature feature extraction[J]. Journal of Energy Storage, 2022, 50: 104646.
11	XING L K, LIU X, LUO W F, et al. State of health estimation for lithium-ion batteries using IAO–SVR[J]. World Electric Vehicle Journal, 2023, 14(5): 122.
12	WANG S L, FERNANDEZ C, YU C M, et al. A novel charged state prediction method of the lithium ion battery packs based on the composite equivalent modeling and improved splice Kalman filtering algorithm[J]. Journal of Power Sources, 2020, 471: 228450.
13	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.
14	高德欣, 刘欣, 杨清. 基于卷积神经网络与双向长短时融合的锂离子电池剩余使用寿命预测[J]. 信息与控制, 2022, 51(3): 318-329, 360.
	GAO D X, LIU X, YANG Q. Remaining useful life prediction of lithium-ion battery based on CNN and BiLSTM fusion[J]. Information and Control, 2022, 51(3): 318-329, 360.
15	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.
16	LIN Z C, HU H P, LIU W, et al. State of health estimation of lithium-ion batteries based on remaining area capacity[J]. Journal of Energy Storage, 2023, 63: 107078.
17	KONG J Z, YANG F F, ZHANG X, et al. Voltage-temperature health feature extraction to improve prognostics and health management of lithium-ion batteries[J]. Energy, 2021, 223: 120114.
18	JIA J F, LIANG J Y, SHI Y H, et al. SOH and RUL prediction of lithium-ion batteries based on Gaussian process regression with indirect health indicators[J]. Energies, 2020, 13(2): 375.
19	ROMAN D, SAXENA S, ROBU V, et al. Machine learning pipeline for battery state-of-health estimation[J]. Nature Machine Intelligence, 2021, 3: 447-456.
20	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.
21	WU C H, WU F Z, QI T, et al. Fastformer: Additive attention can be all you need[J]. ArXiv e-Prints, 2021: arXiv: .
22	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: 383-391.
23	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. December 4 - 9, 2017, Long Beach, California, USA. ACM, 2017: 6000–6010.
24	徐建民, 郭紫芯, 肖涛, 等.基于正交试验的水平换热器充气模拟实验研究[J]. 化学工程与装备, 2023(9): 10-12, 16.
	XU J M, GUO Z X,XIAO T, et al. Experimental study on air-simulated level heat exchanger based on orthogonal experiment[J]. Chemical Engineering and Equipment, 2023(9): 10-12, 16.

参数	B01	B02	B03	B04
额定容量/Ah	1.1	1.1	1.1	1.1
充电速率/C	5.3-4-1	5.6-4.6-1	5.6-4.3-1	5.6-4.6-1
充电切换SOC值	54%,80%	19%,80%	36%,80%	19%,80%
放电速率/C	4	4	4	4
最大截止电压/V	3.6	3.6	3.6	3.6
最小截止电压/V	2.0	2.0	2.0	2.0
循环次数	1062	1266	1114	1047

实验次数	输入步长L	卷积核尺寸K	头数H	网络深度N	RMSE
1	10	3	4	3	0.00304
2	10	4	6	4	0.00269
3	10	5	8	5	0.00303
4	15	3	6	5	0.00314
5	15	4	8	3	0.00291
6	15	5	4	4	0.00368
7	20	3	8	4	0.00300
8	20	4	4	5	0.00283
9	20	5	6	3	0.00265

因素	输入步长m	卷积核尺寸k	头数h	网络深度n
A₁	0.0029	0.00306	0.00318	0.00286
A₂	0.00324	0.00281	0.00282	0.00312
A₃	0.00282	0.00312	0.00298	0.00300
R	0.00042	0.00031	0.00036	0.00025

参数	模型	B01	B03	B04
	RNN	0.296%	0.325%	0.317%
MAE	CNN	0.204%	0.276%	0.261%
	GRU	0.244%	0.251%	0.298%
	卷积Fastformer	0.084%	0.080%	0.255%

	RNN	0.362%	0.396%	0.360%
RMSE	CNN	0.271%	0.311%	0.409%
	GRU	0.314%	0.362%	0.371%
	卷积Fastformer	0.133%	0.102%	0.291%

[1]	胡大林, 任潘利, 张昌明, 杨明阳, 卢周广. Al-Y-Zr原位共掺杂提高4.53 V钴酸锂正极材料的循环性能[J]. 储能科学与技术, 2024, 13(3): 742-748.
[2]	刘剑, 于立博, 吴振兴, 牟介刚. 基于风冷的锂离子电池充放电设备热特性影响研究[J]. 储能科学与技术, 2024, 13(3): 914-923.
[3]	武美玲, 牛磊, 李世友, 赵冬妮. 正极预锂化添加剂用于锂离子电池的研究进展[J]. 储能科学与技术, 2024, 13(3): 759-769.
[4]	张稚国, 李华清, 王莉, 何向明. 锂离子电池塑料-金属复合集流体的特性及制备研究进展[J]. 储能科学与技术, 2024, 13(3): 749-758.
[5]	孙明明. 有机无机复合锂离子电池固态电解质专利分析[J]. 储能科学与技术, 2024, 13(3): 1096-1105.
[6]	宋元明, 刘亚杰, 金光, 周星, 黄旭程. 锂离子电池/超级电容器混合储能系统能量管理方法综述[J]. 储能科学与技术, 2024, 13(2): 652-668.
[7]	李校磊, 高健, 周伟东, 李泓. COMSOL Multiphysics在锂离子电池中的应用[J]. 储能科学与技术, 2024, 13(2): 546-567.
[8]	彭可, 张志成, 胡有章, 张旭辉, 周稼辉, 李彬. 基于有限元的热力耦合场匣钵运动分析与优化[J]. 储能科学与技术, 2024, 13(2): 634-642.
[9]	雷旗开, 余胤, 彭鹏, 陈满, 金凯强, 王青松. 隔热材料布局方式对280 Ah磷酸铁锂电池热失控传播抑制效果的影响[J]. 储能科学与技术, 2024, 13(2): 495-502.
[10]	李珂, 郝奕帆, 方振华, 王静, 张松通, 祝夏雨, 邱景义, 明海. 高功率化学电源体系发展及军事应用分析[J]. 储能科学与技术, 2024, 13(2): 436-461.
[11]	段双明, 张胜利. 基于自适应多层RLS的锂离子电池参数辨识[J]. 储能科学与技术, 2024, 13(2): 712-720.
[12]	杜文, 王君雷, 徐运飞, 李世龙, 王昆. 火焰喷雾热解法生产锂离子电池高镍三元正极材料的技术经济分析[J]. 储能科学与技术, 2024, 13(1): 345-357.
[13]	梁宏毅, 陈锋, 甘友毅, 邵丹. 动力锂电池三元正极低温性能研究[J]. 储能科学与技术, 2024, 13(1): 293-298.
[14]	肖也, 徐磊, 闫崇, 黄佳琦. 锂电池用参比电极的设计与应用[J]. 储能科学与技术, 2024, 13(1): 82-91.
[15]	戴雪娇, 闫婕, 王管, 董浩天, 蒋丹枫, 魏泽威, 孟凡星, 刘松涛, 张海涛. 铌基低温电池关键材料研究进展[J]. 储能科学与技术, 2024, 13(1): 311-324.