基于卡尔曼滤波算法优化Transformer模型的锂离子电池健康状态预测方法

doi:10.19799/j.cnki.2095-4239.2024.0145

摘要/Abstract

摘要：

回顾了锂离子电池健康状态(state of health, SOH)的预测方法，本文提出了一种基于卡尔曼滤波器（Kalman filter）优化的Transformer网络，利用历史数据预测电池SOH。首先，通过添加高斯噪声、自动编码器重构对电池数据进行预处理，去除电池数据中原始噪声并强化数据特征；其次，利用提出的KF-Transformer（Kalman filter-transformer）算法模型提取电池健康状态变化特征，使得Transformer网络能够更好地捕捉电池健康状态的非线性变化；最后，通过线性层完成电池健康状态变化特征到电池健康状态预测的映射，得到锂离子电池健康状态的预测结果。本文使用3个不同充放电策略、不同测试环境下的锂离子电池数据集[分别为美国国家航空航天局(NASA)数据集、马里兰大学CALCE CS2数据集和CALCE CX2数据集]来验证本文提出的锂离子电池健康状态预测算法的鲁棒性和准确性，且本文算法对于SOH不同状态、不同循环次数的预测均具有较好的结果。研究结果显示，本文方法的平均绝对百分比误差(mean absolute percentage error, MAPE)能控制在2%，决定系数（R-square, R²）为0.987，并与多层感知机(MLP, multilayer perceptron)、循环神经网络(RNN, recurrent neura network)、长短时记忆(LSTM, long short-term memory)、门控循环单元(GRU, gated recurrent unit)进行比较，证明了本文方法的优越性。

关键词: 锂离子电池, 电池健康状态, 自动编码器, Transformer网络, 卡尔曼滤波器

Abstract:

Reviewing methods for predicting the lithium-ion battery state of health(SOH), we propose a battery SOH prediction method, which uses Transformer network based on Kalman Filter. Firstly, the battery data is preprocessed by adding Gaussian noise and auto-encoder reconstrution to remove the original noise from the data and strengthen the data features. Secondly, the battery data is extracted using the proposed (Kalman Filter-Transformer, KF-Transformer) model to extract the features of battery SOH changes, so that the Transformer network can better capture the nonlinear changes of battery SOH. Finally, the mapping from the features to the prediction of battery SOH is accomplished through the linear layer to obtain the prediction results of lithium-ion battery SOH. In this paper, three datasets (NASA, CALCE CS2 and CALCE CX2) are used for training and testing, and the two datasets use different temperatures and different batteries. The R² value of the predicted result in this paper is 0.987 and the mean absolute percentage error (MAPE) value is 1.8%, which is better than Multilayer Perceptron (MLP), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU). These results demonstrate the accuracy and stability of the proposed method for SOH prediction.

Key words: lithium-ion battery, battery state of health, auto-encoder, transformer network, Kalman Filter

中图分类号:

TM 911

黄煜峰, 梁焕超, 许磊. 基于卡尔曼滤波算法优化Transformer模型的锂离子电池健康状态预测方法[J]. 储能科学与技术, 2024, 13(8): 2791-2802.

Yufeng HUANG, Huanchao LIANG, Lei XU. Kalman filter optimize Transformer method for state of health prediction on lithium-ion battery[J]. Energy Storage Science and Technology, 2024, 13(8): 2791-2802.

图/表 16

表1

图1

图2

表2

图3

表3

图4

图5

表4

图6

图7

表5

图8

表6

表7

表8

参考文献 37

1	HANNAN M A, LIPU M S H, HUSSAIN A, et al. A review of lithium-ion battery state of charge estimation and management system in electric vehicle applications: Challenges and recommendations[J]. Renewable and Sustainable Energy Reviews, 2017, 78: 834-854. DOI: 10.1016/j.rser.2017.05.001.
2	SHRIVASTAVA P, SOON T K, BIN IDRIS M Y I, et al. Overview of model-based online state-of-charge estimation using Kalman filter family for lithium-ion batteries[J]. Renewable and Sustainable Energy Reviews, 2019, 113: DOI: 10.1016/j.rser. 2019.06.040.
3	WU L X, LIU K, PANG H. Evaluation and observability analysis of an improved reduced-order electrochemical model for lithium-ion battery[J]. Electrochimica Acta, 2021, 368: DOI: 10.1016/j.electacta.2020.137604.
4	LIU Z Y, ZHAO J J, WANG H, et al. A new lithium-ion battery SOH estimation method based on an indirect enhanced health indicator and support vector regression in PHMs[J]. Energies, 2020, 13(4): DOI: 10.3390/en13040830.
5	HU X S, XU L, LIN X K, et al. Battery lifetime prognostics[J]. Joule, 2020, 4(2): 310-346. DOI: 10.1016/j.joule.2019.11.018.
6	ASHWIN T R, CHUNG Y M, WANG J H. Capacity fade modelling of lithium-ion battery under cyclic loading conditions[J]. Journal of Power Sources, 2016, 328: 586-598. DOI: 10.1016/j.jpowsour. 2016.08.054.
7	MISHRA M, MARTINSSON J, RANTATALO M, et al. Bayesian hierarchical model-based prognostics for lithium-ion batteries[J]. Reliability Engineering & System Safety, 2018, 172: 25-35. DOI: 10.1016/j.ress.2017.11.020.
8	POLA D A, NAVARRETE H F, ORCHARD M E, et al. Particle-filtering-based discharge time prognosis for lithium-ion batteries with a statistical characterization of use profiles[J]. IEEE Transactions on Reliability, 2015, 64(2): 710-720. DOI: 10.1109/TR.2014.2385069.
9	MO B H, YU J S, TANG D Y, et al. A remaining useful life prediction approach for lithium-ion batteries using Kalman filter and an improved particle filter[C]//2016 IEEE International Conference on Prognostics and Health Management (ICPHM). Ottawa, ON, Canada. IEEE, 2016: 1-5. DOI: 10.1109/ICPHM. 2016.7542847.
10	YE L H, CHEN S J, SHI Y F, et al. Remaining useful life prediction of lithium-ion battery based on chaotic particle swarm optimization and particle filter[J]. International Journal of Electrochemical Science, 2023, 18(5): DOI: 10.1016/j.ijoes. 2023.100122.
11	ZHOU Y P, HUANG M H. Lithium-ion batteries remaining useful life prediction based on a mixture of empirical mode decomposition and ARIMA model[J]. Microelectronics Reliability, 2016, 65: 265-273. DOI: 10.1016/j.microrel.2016.07.151.
12	ZHANG Y Z, XIONG R, HE H W, et al. Lithium-ion battery remaining useful life prediction with box-cox transformation and Monte Carlo simulation[J]. IEEE Transactions on Industrial Electronics, 2019, 66(2): 1585-1597. DOI: 10.1109/TIE.2018.2808918.
13	ZHANG Y Z, XIONG R, HE H W, et al. Long short-term memory recurrent neural network for remaining useful life prediction of lithium-ion batteries[J]. IEEE Transactions on Vehicular Technology, 2018, 67(7): 5695-5705. DOI: 10.1109/TVT.2018.2805189.
14	REN L, ZHAO L, HONG S, et al. Remaining useful life prediction for lithium-ion battery: A deep learning approach[J]. IEEE Access, 2018, 6: 50587-50598. DOI: 10.1109/ACCESS.2018.2858856.
15	QIN X L, ZHAO Q, ZHAO H B, et al. Prognostics of remaining useful life for lithium-ion batteries based on a feature vector selection and relevance vector machine approach[C]//2017 IEEE International Conference on Prognostics and Health Management (ICPHM). Dallas, TX, USA. IEEE, 2017: 1-6. DOI: 10.1109/ICPHM.2017.7998297.
16	NUHIC A, TERZIMEHIC T, SOCZKA-GUTH T, et al. Health diagnosis and remaining useful life prognostics of lithium-ion batteries using data-driven methods[J]. Journal of Power Sources, 2013, 239: 680-688. DOI: 10.1016/j.jpowsour.2012.11.146.
17	WENG C H, CUI Y J, SUN J, et al. On-board state of health monitoring of lithium-ion batteries using incremental capacity analysis with support vector regression[J]. Journal of Power Sources, 2013, 235: 36-44. DOI: 10.1016/j.jpowsour.2013.02.012.
18	LVOVICH V, WU J, BENNETT W, et al. Applications of AC impedance spectroscopy as characterization and diagnostic tool in Li-metal battery cells[J]. ECS Transactions, 2014, 58(22): 1-14. DOI: 10.1149/05822.0001ecst.
19	HUANG J, LI Z, ZHANG J B. Dynamic electrochemical impedance spectroscopy reconstructed from continuous impedance measurement of single frequency during charging/discharging[J]. Journal of Power Sources, 2015, 273: 1098-1102. DOI: 10.1016/j.jpowsour.2014.07.067.
20	TRAN M K, MATHEW M, JANHUNEN S, et al. A comprehensive equivalent circuit model for lithium-ion batteries, incorporating the effects of state of health, state of charge, and temperature on model parameters[J]. Journal of Energy Storage, 2021, 43: DOI: 10.1016/j.est.2021.103252.
21	DING X F, ZHANG D H, CHENG J W, et al. An improved Thevenin model of lithium-ion battery with high accuracy for electric vehicles[J]. Applied Energy, 2019, 254: 113615. DOI: 10.1016/j.apenergy.2019.113615.
22	ZHANG X Q, ZHANG W P, LEI G Y. A review of Li-ion battery equivalent circuit models[J]. Transactions on Electrical and Electronic Materials, 2016, 17(6): 311-316. DOI: 10.4313/teem.2016.17.6.311.
23	PLETT G L. High-performance battery-pack power estimation using a dynamic cell model[J]. IEEE Transactions on Vehicular Technology, 2004, 53(5): 1586-1593. DOI: 10.1109/TVT.2004.832408.
24	LI Y, LIU K L, FOLEY A M, et al. Data-driven health estimation and lifetime prediction of lithium-ion batteries: A review[J]. Renewable and Sustainable Energy Reviews, 2019, 113: 109254. DOI: 10.1016/j.rser.2019.109254.
25	VANEM E, SALUCCI C B, BAKDI A, et al. Data-driven state of health modelling—A review of state of the art and reflections on applications for maritime battery systems[J]. Journal of Energy Storage, 2021, 43: 103158. DOI: 10.1016/j.est.2021.103158.
26	YANG D D, LU J Y, DONG H L, et al. Pipeline signal feature extraction method based on multi-feature entropy fusion and local linear embedding[J]. Systems Science & Control Engineering, 2022, 10(1): 407-416. DOI: 10.1080/21642583. 2022.2063202.
27	LIANG C M, LI Y W, LIU Y H, et al. Segmentation and weight prediction of grape ear based on SFNet-ResNet18[J]. Systems Science & Control Engineering, 2022, 10(1): 722-732. DOI: 10.1080/21642583.2022.2110541.
28	LI X G, FENG S, HOU N, et al. Surface microseismic data denoising based on sparse autoencoder and Kalman filter[J]. Systems Science & Control Engineering, 2022, 10(1): 616-628. DOI: 10.1080/21642583.2022.2087786.
29	FINE T L. Fundamentals of artificial neural networks[book reviews[J]. IEEE Transactions on Information Theory, 1996, 42(4): 1322. DOI: 10.1109/TIT.1996.508868.
30	WANG Z P, MA J, ZHANG L. State-of-health estimation for lithium-ion batteries based on the multi-island genetic algorithm and the Gaussian process regression[J]. IEEE Access, 2017, 5: 21286-21295. DOI: 10.1109/ACCESS.2017.2759094.
31	SHE C Q, LI Y, ZOU C F, et al. Offline and online blended machine learning for lithium-ion battery health state estimation[J]. IEEE Transactions on Transportation Electrification, 2022, 8(2): 1604-1618. DOI: 10.1109/TTE.2021.3129479.
32	陈欣, 李云伍, 梁新成, 等. 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. DOI: 10.19799/j.cnki.2095-4239.2023.0323.
	CHEN X, LI Y W, LIANG X C, et al. Battery health state estimation of combined Transformer-GRU based on modal decomposition[J]. Energy Storage Science and Technology, 2023, 12(9): 2927-2936. DOI: 10.19799/j.cnki.2095-4239. 2023.0323.
33	毛百海, 覃吴, 肖显斌, 等. 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. DOI: 10.19799/j.cnki.2095-4239.2023.0514.
	MAO B H, QIN W, XIAO X B, et al. SOH estimation of lithium-ion batteries based on LSTM & GRU-Attention multijoint model[J]. Energy Storage Science and Technology, 2023, 12(11): 3519-3527. DOI: 10.19799/j.cnki.2095-4239.2023.0514.
34	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. DOI: 10.19799/j.cnki.2095-4239.2023.0735.
	SHEN X Y, YIN C B. SOH estimation of lithium-ion batteries using a convolutional Fastformer[J]. Energy Storage Science and Technology, 2024, 13(3): 990-999. DOI: 10.19799/j.cnki.2095-4239.2023.0735.
35	HE W, WILLIARD N, OSTERMAN M, et al. Prognostics of lithium-ion batteries based on Dempster-shafer theory and the Bayesian Monte Carlo method[J]. Journal of Power Sources, 2011, 196(23): 10314-10321. DOI: 10.1016/j.jpowsour.2011.08.040.
36	XING Y J, MA E W M, TSUI K L, et al. An ensemble model for predicting the remaining useful performance of lithium-ion batteries[J]. Microelectronics Reliability, 2013, 53(6): 811-820. DOI: 10.1016/j.microrel.2012.12.003.
37	WILLIARD N, HE W, OSTERMAN M, et al. Comparative analysis of features for determining state of health in lithium-ion batteries[J]. International Journal of Prognostics and Health Management, 2020, 4(1): DOI: 10.36001/ijphm.2013.v4i1.1437.

参数	NASA	CALCE CS2	CALCE CX2
电池容量	2000 mAh	1100 mAh	1350 mAh
电芯材质	LiNiMnCo/石墨	LiCoO₂	LiCoO₂
标称电压	3.7 V	3.7 V	3.7 V
测试温度	24 ℃	1 ℃	0.5 ℃
充/放电截止电压	4.2V/2.5 V	4.2V/2.7 V	4.2V/2.7 V
充/放电速率	2 A(1C)	1.1 A(1C)	0.675 A(0.5C)
电池编号	B0005, B0006, B0007, B0018	CS2-35, CS2-36, CS2-37, CS2-38	CX2-34, CX2-36, CX2-37, CX2-38

参数名称	参数量
epoch	5000
学习率	0.0005
注意力数量	16
隐藏层维度	16
隐藏层数量	8
注意力窗口大小	64

Index	RNN	MLP	LSTM	GRU	Ours
Index	RNN	MLP	LSTM	GRU	Before-Transformer	In-Transformer	After-Transformer
R₂	0.8214	0.8047	0.8785	0.8592	0.9627	0.9789	0.9612
RMSE	0.0911	0.0959	0.0966	0.0988	0.0497	0.0345	0.0448
MAPE	0.0540	0.0564	0.0568	0.0575	0.0385	0.0268	0.0341

Index	RNN	MLP	LSTM	GRU	Ours
Index	RNN	MLP	LSTM	GRU	Before-Transformer	In-Transformer	After-Transformer
R²	0.8063	0.8384	0.8551	0.8496	0.9765	0.9874	0.9542
RMSE	0.0992	0.1115	0.0684	0.0632	0.0305	0.0252	0.0531
MAPE	0.1211	0.1554	0.0913	0.0841	0.0237	0.0189	0.0415

Index	RNN	MLP	LSTM	GRU	Ours
Index	RNN	MLP	LSTM	GRU	Before-Transformer	In-Transformer	After-Transformer
R²	0.8973	0.9200	0.9150	0.9031	0.9674	0.9803	0.9587
RMSE	0.0354	0.0339	0.0349	0.0356	0.0268	0.0171	0.0281
MAPE	0.0336	0.0325	0.0334	0.0841	0.0215	0.0137	0.0269