融合特征多尺度池化的锂电池健康状态评估算法

周星凯; 竺春祥; 贾迎浩; 王斌锐; 李冲

doi:10.19799/j.cnki.2095-4239.2026.0153

您当前的位置：

首页 >

文章列表页 >

融合特征多尺度池化的锂电池健康状态评估算法

储能XXXX | 更新时间：2026-04-29

- 融合特征多尺度池化的锂电池健康状态评估算法
- Multi-Scale Pooling Enhanced Feature Fusion for Lithium-Ion Battery SOH Estimation
- 储能科学与技术 2026年页码：1-15
- 作者机构：
  
  1.中国计量大学机电工程学院，杭州 310018
  2.上海第二工业大学智能制造与控制工程学院，上海 201209
- 作者简介：
  
  周星凯（2000—），男，硕士生，锂电池健康状态估计，E-mail：1197907916@qq.com ；
  竺春祥，博士，讲师，研究方向，锂电池管理系统算法。，E-mail：led2016@cjlu.edu.cn 。
- 基金信息：
  
  国家自然科学基金资助项目(52205292);浙江省自然科学基金白马湖实验室联合基金资助项目(LBMHY25B030001);中国计量大学2025年校级教改项目(HEX2025003);中国计量大学2024年度基本科研业务费项目(2024YW59)
- DOI：10.19799/j.cnki.2095-4239.2026.0153
  中图分类号： TM911
- 收稿：2026-02-10，
  
  修回：2026-04-27，
  
  网络首发：2026-04-28，
- 稿件说明：
移动端阅览
周星凯, 竺春祥, 贾迎浩, 等. 融合特征多尺度池化的锂电池健康状态评估算法[J]. 储能科学与技术, XXXX, XX(XX): 1-15.

ZHOU Xingkai, ZHU Chunxiang, JIA Yinghao, et al. Multi-Scale Pooling Enhanced Feature Fusion for Lithium-Ion Battery SOH Estimation[J]. Energy Storage Science and Technology, XXXX, XX(XX): 1-15.
周星凯, 竺春祥, 贾迎浩, 等. 融合特征多尺度池化的锂电池健康状态评估算法[J]. 储能科学与技术, XXXX, XX(XX): 1-15. DOI： 10.19799/j.cnki.2095-4239.2026.0153.

ZHOU Xingkai, ZHU Chunxiang, JIA Yinghao, et al. Multi-Scale Pooling Enhanced Feature Fusion for Lithium-Ion Battery SOH Estimation[J]. Energy Storage Science and Technology, XXXX, XX(XX): 1-15. DOI： 10.19799/j.cnki.2095-4239.2026.0153.

摘要

随着新能源技术的迅速发展，准确评估电池健康状态对于电池管理系统的优化和电池寿命预测具有至关重要的意义。现有的SOH估计方法主要依赖传统神经网络对时间序列进行编码，但这些方法通常忽略了通道间特征的相互关系，导致信息利用不足。为解决这一问题，本文提出了一种基于混合空间池化与多通道特征加权融合的SOH预测方法。首先，通过混合池化模块对电池电压、电流、温度等时序数据进行多尺度特征提取，结合最大池化、平均池化和L2池化策略，有效提高了模型对退化特征的感知能力。接着，利用多通道加权融合模块，通过优化特征通道之间的交互与融合，进一步提升了模型的预测性能。实验结果表明，所提方法在马里兰大学、同济大学和西交大的多个公开电池数据集上表现优异，在MAE、RMSE和MAPE等评估指标上均优于传统方法，取得了最佳结果，其中MAE为0.0012，RMSE为0.0017，MAPE为0.1661。尤其在复杂退化模式下，展现出更强的鲁棒性和更好的泛化能力。

Abstract

Accurate estimation of lithium-ion battery State of Health (SOH) is essential for the optimization of battery management systems (BMS) and reliable lifetime prediction. However

existing neural-network-based SOH estimation methods often ignore the interdependencies among feature channels

leading to suboptimal feature utilization and degraded prediction accuracy.This study proposes a novel SOH estimation framework that integrates Hybrid Spatial Pooling (HPP) with a multi-channel feature fusion module (STAR). The HPP module performs multi-scale feature extraction on voltage

current

and temperature time-series by combining max

average

and L2 pooling operations to enhance degradation-sensitive feature representation. The STAR module further strengthens feature interactions across channels through adaptive weighting and cross-channel enhancement.Experimental evaluations on three publicly available datasets from the University of Maryland

Tongji University

and Xi'an Jiaotong University demonstrate the effectiveness of the proposed approach. The model achieves a mean absolute error of approximately 0.0012

a root-mean-square error of roughly 0.0017

and a mean absolute percentage error of about 0.1661

consistently outperforming conventional neural-network baselines. In addition

the method exhibits strong robustness under complex degradation patterns and preserves stable generalization capability across different battery chemistries.The proposed HPP–STAR framework significantly enhances SOH estimation through three main contributions: (1) introducing hybrid multi-scale pooling to enrich degradation-aware representations

(2) optimizing inter-channel dependencies via multi-channel feature fusion

and (3) achieving consistently high accuracy across diverse datasets. These results verify the effectiveness and generalization capability of the proposed approach

providing a promising direction for future SOH estimation in practical BMS applications.

关键词

Keywords

references

SHU X, SHEN J, GUO F, et al. Towards real-world battery health intelligence: A review of machine learning advances and challenges in SOH estimation[J]. eTransportation, 2025, 26: 100509. https://doi.org/10.1016/j.etran.2025.100509

S V, CHE H S, SELVARAJ J, et al. State of health (SOH) estimation methods for second life lithium-ion battery—Review and challenges[J]. Applied Energy, 2024, 369: 123542. https://doi.org/10.1016/j.apenergy.2024.123542.

ALSUWIAN T, ANSARI S, ZAINURI M A A M, et al. A review of expert hybrid and co-estimation techniques for SOH and RUL estimation in battery management system with electric vehicle application[J]. Expert Systems with Applications, 2024, 246: 123123. https://doi.org/10.1016/j.eswa.2023.123123.

ZHANG L, CHEN C, LUO D. A comprehensive framework of synchronous SOC-SOH joint estimation for lithium-ion battery with multi-depth expert networks[J]. Energy, 2025, 339: 138953. https://doi.org/10.1016/j.energy.2025.138953.

YANG S, ZHANG C, JIANG J, et al. Review on state-of-health of lithium-ion batteries: Characterizations, estimations and applications[J]. Journal of Cleaner Production, 2021, 314: 128015. https://doi.org/10.1016/j.jclepro.2021.128015.

WANG R, CHEN L, XU J, et al. Multimodal deep learning with time-frequency health features for battery SOH and RUL prediction[J]. Journal of Energy Chemistry, 2026, 113: 303-314. https://doi.org/10.1016/j.jechem.2025.09.018.

DEMIRCI O, TASKIN S, SCHALTZ E, et al. Review of battery state estimation methods for electric vehicles—Part II: SOH estimation[J]. Journal of Energy Storage, 2024, 96: 112703. https://doi.org/10.1016/j.est.2024.112703.

CHEN G, ZHOU H, XU Y, et al. A novel SOC-OCV separation and extraction technology suitable for online joint estimation of SOC and SOH in lithium-ion batteries[J]. Energy, 2025, 326: 136246. https://doi.org/10.1016/j.energy.2025.136246.

Raza K, Samadi A F, Berecibar M, 等. From EV to stationary energy storage: EIS-based SoH estimation for second life Li-ion batteries[J]. Journal of Energy Storage, 2026, 141: 119316. DOI: https://doi.org/10.1016/j.est.2025.119316.

QIN D, LI L, WU H, et al. A composite progressive approach alternating between electrochemical model and data-driven model for the state of health assessment of lithium-ion batteries[J]. Journal of Energy Storage, 2024, 97: 112921. https://doi.org/10.1016/j.est.2024.112921.

SUN J, LIU Y, KAINZ J. A comparative study of parameter identification methods for equivalent circuit models for lithium-ion batteries and their application to state of health estimation[J]. Journal of Energy Storage, 2025, 114: 115707. https://doi.org/10.1016/j.est.2025.115707.

BAO Z, NIE J, LIN H, et al. Dual-task learning for joint state-of-charge and state-of-energy estimation of lithium-ion battery in electric vehicle[J]. IEEE Transactions on Transportation Electrification, 2025, 11(1): 558-569. https://doi.org/10.1109/TTE.2024.3393477.

ZHANG J, WANG P, GONG Q, et al. SOH estimation of lithium-ion batteries based on least squares support vector machine error compensation model[J]. Journal of Power Electronics, 2021, 21(11): 1712-1723. https://doi.org/10.1007/s43236-021-00307-8.

YANG D, ZHANG X, PAN R, et al. A novel Gaussian process regression model for state-of-health estimation of lithium-ion battery using charging curve[J]. Journal of Power Sources, 2018, 384: 387-395. https://doi.org/10.1016/j.jpowsour.2018.03.015.

YANG N, SONG Z, 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. https://doi.org/10.1016/j.est.2021.103857.

LIU J F, CHEN H W, IU H H C. Lithium-ion battery life prediction considering capacity regeneration using VMD and DAIPSO-GPR[J]. Journal of Electronics & Information Technology, 2023, 45(3): 1111-1120. https://doi.org/10.11999/JEIT211585.刘金锋, 陈浩伟, IU H H C. 基于VMD和DAIPSO-GPR解决容量再生现象的锂离子电池寿命预测研究[J]. 电子与信息学报, 2023, 45(3): 1111-1120.

ZHANG L, JI T, YU S, et al. Accurate prediction approach of SOH for lithium-ion batteries based on LSTM method[J]. Batteries, 2023, 9(3): 177. https://doi.org/10.3390/batteries9030177.

ZHU C, GAO M, HE Z, et al. State of health prediction for Li-ion batteries with end-to-end deep learning[J]. Journal of Energy Storage, 2023, 65: 107218. https://doi.org/10.1016/j.est.2023.107218.

GUO M H, LIU Z N, MU T J, et al. Beyond self-attention: External attention using two linear layers for visual tasks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(5): 5436-5447. https://doi.org/10.1109/TPAMI.2022.3211006.

WANG Q, WU B, ZHU P, et al. ECA-Net: Efficient channel attention for deep convolutional neural networks[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 11531-11539.

TAN C, GAO Z, WU L, et al. Temporal attention unit: Towards efficient spatiotemporal predictive learning[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023: 18770-18782.

VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C ] //Advances in Neural Information Processing Systems. 2017, 30 .

BAO Z, NIE J, LIN H, et al. TTSNet: State-of-charge estimation of Li-ion battery in electrical vehicles with temporal transformer-based sequence network[J]. IEEE Transactions on Vehicular Technology, 2024, 73(6): 7838-7851. https://doi.org/10.1109/TVT.2024.3350663.

WANG C, BAO Z, LIN H, et al. A multi-task targeted learning framework for lithium-ion battery state-of-health and remaining useful life[J]. IEEE Transactions on Transportation Electrification, 2025: 1-1. https://doi.org/10.1109/TTE.2025.3610457.

GUI X, DU J, WANG Q, et al. Multi-modal data information alignment based SOH estimation for lithium-ion batteries using a local–global parallel CNN-Transformer network[J]. Journal of Energy Storage, 2025, 129: 117178. https://doi.org/10.1016/j.est.2025.117178.

HE K, ZHANG X, REN S, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[C]//Computer Vision – ECCV 2014. Cham: Springer, 2014: 346-361.

HAN L, CHEN X Y, YE H J, et al. Softs: Efficient multivariate time series forecasting with series-core fusion[J]. Advances in Neural Information Processing Systems, 2024, 37: 64145-64175.

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. https://doi.org/10.1016/j.jpowsour.2011.08.040.

WANG F, ZHAI Z, ZHAO Z, et al. Physics-informed neural network for lithium-ion battery degradation stable modeling and prognosis[J]. Nature Communications, 2024, 15(1): 4332. https://doi.org/10.1038/s41467-024-48779-z.

ZHU J, WANG Y, HUANG Y, et al. Data-driven capacity estimation of commercial lithium-ion batteries from voltage relaxation[J]. Nature Communications, 2022, 13(1): 2261. https://doi.org/10.1038/s41467-022-29837-w.

LI Y, QIN X, CHAI M, et al. SOH evaluation and RUL estimation of lithium-ion batteries based on MC-CNN-TimesNet model[J]. Reliability Engineering & System Safety, 2025, 261: 111125. https://doi.org/10.1016/j.ress.2025.111125.

CHEN B, ZHANG Y, WU J, et al. Lithium-ion battery state of health estimation based on feature reconstruction and transformer-GRU parallel architecture[J]. Energies, 2025, 18(5). https://doi.org/10.3390/en18051236.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于深度特征融合模型的电动汽车不同模式充电时长自适应预测

基于白鹭群优化高斯过程回归的锂电池SOH估计方法

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

基于自适应最优组合核函数高斯过程回归的锂电池健康状态区间估计

考虑锂电池温度和老化的荷电状态估算