一种并行多尺度特征融合模型开展的基于弛豫电压的锂电池SOH估计研究

doi:10.19799/j.cnki.2095-4239.2024.0808

摘要/Abstract

摘要：

锂离子电池健康状态(state of health, SOH)估计对确保能量存储系统的可靠性和安全性至关重要。然而，现有SOH估计方法在单一特征提取和固定充放电条件依赖方面存在局限性，难以适应多变的实际工作环境。为解决这一问题，本工作提出了一种基于弛豫电压的并行多尺度特征融合卷积模型(multi-scale feature fusion convolution model, MSFFCM)结合极端梯度提升树(XGBoost)的SOH估计方法。MSFFCM通过多层堆叠卷积模块提取弛豫电压数据的深层特征，同时利用并行多尺度注意力机制增强了多尺度特征的捕捉能力，并将这些特征与统计特征进行融合，以提升模型的特征提取和融合能力。针对XGBoost模型，本工作应用贝叶斯优化算法进行参数调优，从而在多源融合特征基础上实现高精度SOH估计。实验验证基于两种商用18650型号电池的多温度和多充放电策略数据集，结果表明该方法的均方根误差(RMSE)和平均绝对误差(MAE)均小于0.5%，明显优于传统方法。本工作为锂电池健康管理提供了一种不依赖特定充放电条件的有效估计工具，有望在复杂的实际应用中发挥重要作用。

关键词: 锂离子电池, 健康状态估计, 弛豫电压, 并行多尺度特征, 特征融合

Abstract:

The estimation of state-of-health (SOH) in lithium-ion batteries is crucial for ensuring the reliability and safety of energy storage systems. However, existing SOH estimation methods are limited by single-feature extraction and dependency on fixed charge-discharge conditions, which hinders adaptability to dynamic operational environments. To address these challenges, we propose an SOH estimation approach based on relaxation voltage, integrating a parallel multiscale feature fusion convolution model (MSFFCM) with XGBoost. The MSFFCM model leverages multilayer stacked convolutional modules to extract deep features from relaxation voltage data while utilizing a parallel multiscale attention mechanism to enhance the capture of multiscale features. These features are then fused with statistical features to improve the model's feature extraction and integration capabilities. Bayesian optimization is applied to the XGBoost model for parameter tuning, enabling high-accuracy SOH estimation based on multi-source fused features. Validation experiments were conducted on datasets from two commercial 18650 lithium-ion battery types under varied temperature and charge-discharge strategies. Results indicate that the proposed method achieves a root mean square error and mean absolute error of less than 0.5%, significantly outperforming conventional methods. This study provides an effective estimation tool for lithium-ion battery health management without reliance on specific charge-discharge conditions, demonstrating promising potential for complex real-world applications.

Key words: lithium-ion battery, state-of-health estimation, relaxation voltage, parallel multi-scale features, feature fusion

中图分类号:

TM 912

王海瑞, 徐长宇, 朱贵富, 侯晓建. 一种并行多尺度特征融合模型开展的基于弛豫电压的锂电池SOH估计研究[J]. 储能科学与技术, 2025, 14(2): 799-811.

Hairui WANG, Changyu XU, Guifu ZHU, Xiaojian HOU. A parallel multi cale-featured fusion model for state-of-health estimation of lithium-ion batteries based on relaxation voltage[J]. Energy Storage Science and Technology, 2025, 14(2): 799-811.

图/表 16

图1

表1

图3

图4

图5

图6

表2

表3

表4

表5

表6

图7

图8

图9

图10

图11

参考文献 31

1	尹杰, 刘博, 孙国兵, 等. 基于迁移学习和降噪自编码器-长短时间记忆的锂离子电池剩余寿命预测[J]. 电工技术学报, 2024, 39(1): 289-302. DOI: 10.19595/j.cnki.1000-6753.tces.221890.
	YIN J, LIU B, SUN G B, et al. Transfer learning denoising autoencoder-long short term memory for remaining useful life prediction of Li-ion batteries[J]. Transactions of China Electrotechnical Society, 2024, 39(1): 289-302. DOI: 10.19595/j.cnki.1000-6753.tces.221890.
2	黄凯, 丁恒, 郭永芳, 等. 基于数据预处理和长短期记忆神经网络的锂离子电池寿命预测[J]. 电工技术学报, 2022, 37(15): 3753-3766. DOI: 10.19595/j.cnki.1000-6753.tces.210860.
	HUANG K, DING H, GUO Y F, et al. Prediction of remaining useful life of lithium-ion battery based on adaptive data preprocessing and long short-term memory network[J]. Transactions of China Electrotechnical Society, 2022, 37(15): 3753-3766. DOI: 10.19595/j.cnki.1000-6753.tces.210860.
3	李英顺, 阚宏达, 郭占男, 等. 基于数据预处理和VMD-LSTM-GPR的锂离子电池剩余寿命预测[J]. 电工技术学报, 2024, 39(10): 3244-3258. DOI: 10.19595/j.cnki.1000-6753.tces.230210.
	LI Y S, KAN H D, GUO Z N, et al. Prediction of remaining useful life of lithium-ion battery based on data preprocessing and VMD-LSTM-GPR[J]. Transactions of China Electrotechnical Society, 2024, 39(10): 3244-3258. DOI: 10.19595/j.cnki.1000-6753.tces.230210.
4	FAN Y X, XIAO F, LI C R, et al. A novel deep learning framework for state of health estimation of lithium-ion battery[J]. Journal of Energy Storage, 2020, 32: 101741. DOI:10.1016/j.est.2020.101741.
5	TANG A H, HUANG Y K, XU Y C, et al. Data-physics-driven estimation of battery state of charge and capacity[J]. Energy, 2024, 294: 130776. DOI:10.1016/j.energy.2024.130776.
6	SUI X, HE S, VILSEN S B, et al. A review of non-probabilistic machine learning-based state of health estimation techniques for Lithium-ion battery[J]. Applied Energy, 2021, 300: 117346. DOI:10.1016/j.apenergy.2021.117346.
7	REN Z, DU C Q. A review of machine learning state-of-charge and state-of-health estimation algorithms for lithium-ion batteries[J]. Energy Reports, 2023, 9: 2993-3021. DOI:10.1016/j.egyr.2023.01.108.
8	VILSEN S B, STROE D I. Battery state-of-health modelling by multiple linear regression[J]. Journal of Cleaner Production, 2021, 290: 125700. DOI:10.1016/j.jclepro.2020.125700.
9	SUI X, HE S, MENG J H, et al. Fuzzy entropy-based state of health estimation for Li-ion batteries[J]. IEEE Journal of Emerging and Selected Topics in Power Electronics, 2021, 9(4): 5125-5137. DOI:10.1109/JESTPE.2020.3047004.
10	LIN M Q, WU D G, MENG J H, et al. A multi-feature-based multi-model fusion method for state of health estimation of lithium-ion batteries[J]. Journal of Power Sources, 2022, 518: 230774. DOI:10.1016/j.jpowsour.2021.230774.
11	TAO J, WANG S, CAO W, et al. A comprehensive review of multiple physical and data-driven model fusion methods for accurate lithium-ion battery inner state factor estimation[J]. Batteries, 2024, 10(12): 442.
12	HU X S, CHE Y H, LIN X K, et al. Health prognosis for electric vehicle battery packs: A data-driven approach[J]. IEEE/ASME Transactions on Mechatronics, 2020, 25(6): 2622-2632. DOI:10.1109/TMECH.2020.2986364.
13	WU J, LIU X T, MENG J H, et al. Cloud-to-edge based state of health estimation method for Lithium-ion battery in distributed energy storage system[J]. Journal of Energy Storage, 2021, 41: 102974. DOI:10.1016/j.est.2021.102974.
14	MA Y, SHAN C, GAO J W, et al. A novel method for state of health estimation of lithium-ion batteries based on improved LSTM and health indicators extraction[J]. Energy, 2022, 251: 123973. DOI:10.1016/j.energy.2022.123973.
15	TANG A, JIANG Y, YU Q, et al. A hybrid neural network model with attention mechanism for state of health estimation of lithium-ion batteries[J]. Journal of Energy Storage, 2023, 68: 107734.
16	HU J, WENG L L, GAO Z W, et al. State of health estimation and remaining useful life prediction of electric vehicles based on real-world driving and charging data[J]. IEEE Transactions on Vehicular Technology, 2023, 72(1): 382-394. DOI:10.1109/TVT.2022.3203013.
17	ZHOU K Q, QIN Y, YUEN C. Transfer-learning-based state-of-health estimation for lithium-ion battery with cycle synchronization[J]. IEEE/ASME Transactions on Mechatronics, 2023, 28(2): 692-702. DOI:10.1109/TMECH.2022.3201010.
18	SHEN S, SADOUGHI M, CHEN X Y, et al. A deep learning method for online capacity estimation of lithium-ion batteries[J]. Journal of Energy Storage, 2019, 25: 100817. DOI:10.1016/j.est.2019.100817.
19	YANG N K, SONG Z Y, HOFMANN H, et al. Robust State of Health estimation of lithium-ion batteries using convolutional neural network and random forest[J]. Journal of Energy Storage, 2022, 48: 103857. DOI:10.1016/j.est.2021.103857.
20	WANG S L, TAKYI-ANINAKWA P, JIN S Y, et al. An improved feedforward-long short-term memory modeling method for the whole-life-cycle state of charge prediction of lithium-ion batteries considering current-voltage-temperature variation[J]. Energy, 2022, 254: 124224. DOI:10.1016/j.energy.2022.124224.
21	WANG S L, GAO H Y, TAKYI-ANINAKWA P, et al. Improved multiple feature-electrochemical thermal coupling modeling of lithium-ion batteries at low-temperature with real-time coefficient correction[J]. Protection and Control of Modern Power Systems, 2024, 9(3): 157-173. DOI:10.23919/PCMP.2023.000257.
22	SHEN S Q, LIU B C, ZHANG K, et al. Toward fast and accurate SOH prediction for lithium-ion batteries[J]. IEEE Transactions on Energy Conversion, 2021, 36(3): 2036-2046. DOI:10.1109/TEC.2021.3052504.
23	SU L Z, XU Y, DONG Z Y. State-of-health estimation of lithium-ion batteries: A comprehensive literature review from cell to pack levels[J]. Energy Conversion and Economics, 2024, 5(4): 224-242. DOI:10.1049/enc2.12125.
24	ZHANG M, YIN J, CHEN W L. SOH estimation and RUL prediction of lithium batteries based on multidomain feature fusion and CatBoost model[J]. Energy Science & Engineering, 2023, 11(9): 3082-3101. DOI:10.1002/ese3.1506.
25	HUANG H Y, MENG J H, WANG Y H, et al. A comprehensively optimized lithium-ion battery state-of-health estimator based on Local Coulomb Counting Curve[J]. Applied Energy, 2022, 322: 119469. DOI:10.1016/j.apenergy.2022.119469.
26	SHU X, SHEN J W, LI G, et al. A flexible state-of-health prediction scheme for lithium-ion battery packs with long short-term memory network and transfer learning[J]. IEEE Transactions on Transportation Electrification, 2021, 7(4): 2238-2248. DOI:10.1109/TTE.2021.3074638.
27	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(1): 2261. DOI:10.1038/s41467-022-29837-w.
28	ZHU X F, WANG W, ZOU G P, et al. State of health estimation of lithium-ion battery by removing model redundancy through aging mechanism[J]. Journal of Energy Storage, 2022, 52: 105018. DOI:10.1016/j.est.2022.105018.
29	YANG J F, XIA B, HUANG W X, et al. Online state-of-health estimation for lithium-ion batteries using constant-voltage charging current analysis[J]. Applied Energy, 2018, 212: 1589-1600. DOI:10.1016/j.apenergy.2018.01.010.
30	ZHANG S W, ZHU H P, WU J, et al. Voltage relaxation-based state-of-health estimation of lithium-ion batteries using convolutional neural networks and transfer learning[J]. Journal of Energy Storage, 2023, 73: 108579. DOI:10.1016/j.est. 2023.108579.
31	ZHAO G X, SUN X, XU J J, et al. MUSE: Parallel multi-scale attention for sequence to sequence learning[EB/OL]. 2019: 1911.09483. https://arxiv.org/abs/1911.09483v1

数据集	电池类型	测试温度/℃	充电率/放电率/C	电池数
数据集1	NCA电池	25	0.25/1	7
	型号18650	25	0.5/1	19
	截止电压2.65~4.2 V	25	1/1	9
	额定容量3.5 Ah	35	0.5/1	3
	额定容量3.5 Ah	45	0.5/1	28

数据集2	NCM电池	25	0.5/1	23
	型号18650	35	0.5/1	4
	截止电压2.5~4.2 V	45	0.5/1	28
	额定容量3.5 Ah

参数	数值	参数	数值
Learning_rate	0.08594	Subsample	0.9434
Max_depth	10	Alpha	0.1811
n_estimators	225	Lambda	57.3173

数据集	方法	RMSE/%	MAPE/%	R²
数据集1	未经过优化	0.7315	0.6589	0.9758
数据集1	贝叶斯优化	0.4183	0.368	0.9950

数据集2	未经过优化	0.9274	0.8484	0.9661
数据集2	贝叶斯优化	0.6096	0.5514	0.9885

数据集	方法	RMSE/%	MAPE/%	R²
数据集1	M1	0.8817	0.7835	0.9778
	M2	0.8876	0.8116	0.9775
	M3	0.8273	0.7449	0.9804
	M4	2.016	1.699	0.8839
	M5	0.4183	0.368	0.9950

数据集2	M1	1.007	0.9173	0.9688
	M2	0.9681	0.8829	0.9712
	M3	0.9454	0.8726	0.9725
	M4	1.290	1.131	0.9488
	M5	0.6096	0.5514	0.9885

CNN-BOA	LSTM	Transformer	MSFFCM-XGB
Learning rate:0.0001	Learning rate:2×10^-3	Feature_size:14	Learning rate:0.08
Layer:4	Layer:2	Num_heads:2	Max_depth:10
Dropout:0.001	N_hidden:32	Num_layer:2	N_estimators:225
Batch size:448	Batch size:448	Dropout:0.1	Alphas:0.18
N_epochs:150	N_epochs:200	N_epochs:140	Subsample:0.94