Prediction of service battery capacity considering the characteristics of the degradation phase

ZHOU Tao; MIAO Shuwei

doi:10.19799/j.cnki.2095-4239.2025.0925

您当前的位置：

首页 >

文章列表页 >

Prediction of service battery capacity considering the characteristics of the degradation phase

Energy Storage Test\: Methods and Evaluation | 更新时间：2026-07-01

- Prediction of service battery capacity considering the characteristics of the degradation phase
- Energy Storage Science and Technology Vol. 15, Issue 4, Pages: 1451-1462(2026)
- 作者机构：
  
  三峡大学电气与新能源学院，湖北宜昌 443002
- 作者简介：
- 基金信息：
- DOI：10.19799/j.cnki.2095-4239.2025.0925
  CLC： TM 912;TP 183
- Received：20 October 2025，
  
  Revised：2025-11-17，
  
  Published：28 April 2026
- 稿件说明：
移动端阅览
周涛, 缪书唯. 计及退化阶段特征的服役电池容量预测[J]. 储能科学与技术, 2026, 15(4): 1451-1462.

ZHOU Tao, MIAO Shuwei. Prediction of service battery capacity considering the characteristics of the degradation phase[J]. Energy Storage Science and Technology, 2026, 15(4): 1451-1462.
周涛, 缪书唯. 计及退化阶段特征的服役电池容量预测[J]. 储能科学与技术, 2026, 15(4): 1451-1462. DOI： 10.19799/j.cnki.2095-4239.2025.0925.

ZHOU Tao, MIAO Shuwei. Prediction of service battery capacity considering the characteristics of the degradation phase[J]. Energy Storage Science and Technology, 2026, 15(4): 1451-1462. DOI： 10.19799/j.cnki.2095-4239.2025.0925.

摘要

精准预测锂电池容量对保证其安全稳定运行极为重要。为此，本研究将锂电池简称为电池，将已知全生命周期数据的电池称为测试电池，正在使用中且数据有限的电池称为服役电池，提出基于迁移学习且计及退化阶段特征的服役电池容量预测模型。首先，采用双Bacon-Watts方法，将测试电池全生命周期划分为早期、中期、末期3组退化阶段。接着，构建电池阶段性匹配机制，该机制根据测试电池早期容量数据与服役电池容量数据间的时间扭曲编辑距离，通过两步筛选获得与服役电池适配的测试电池。在退化阶段划分的基础上，通过识别测试电池当前阶段归属以及阶段内的相对位置信息，刻画其阶段编码。随后，应用多层感知机对阶段编码进行特征映射，将映射后的阶段特征嵌入长短期记忆网络中，同时引入加速退化损失项，引导模型学习电池实际退化规律，实现对测试电池的容量预测。最后，将测试电池的预测模型参数进行微调，并迁移至服役电池容量预测任务中。利用麻省理工学院公开数据集进行验证，模型预测平均绝对误差、平均绝对百分比误差及均方根误差均低于1%，为服役电池早期容量预测提供了可靠方案。

Abstract

Accurate prediction of lithium-ion battery capacity is of great significance for its safe and stable operation. Hence

this study refers to lithium-ion batteries as batteries

refers to batteries with known full-life cycle data as test batteries

refers to batteries currently in use with limited data as in-service batteries

and proposes an in-service battery capacity prediction model based on transfer learning that accounts for degradation phase characteristics. First

the double Bacon-Watts method is adopted to divide the entire life cycle of test batteries into three degradation phases: early

middle

and end phases. Subsequently

a phased matching mechanism for batteries is constructed. Based on the Time Warping Edit Distance between the early-stage capacity data of test batteries and the capacity data of in-service batteries

this mechanism obtains test batteries compatible with in-service batteries through a two-step screening process

providing high-quality data samples for subsequent model training. On the basis of phase division

the phase code is characterized by identifying both the current phase attribution of the test battery and the relative position information within that phase. Then

a multi-layer perceptron is applied to perform feature mapping on the phase codes. The mapped phase features are embedded into a long short-term memory network

while an accelerated degradation loss term is introduced simultaneously to guide the model in learning the actual degradation law of batteries

thereby achieving capacity prediction for the test batteries. Finally

the prediction model parameters of the test batteries are fine-tuned and then transferred to the capacity prediction task of in-service batteries. Validation was conducted using the public dataset from the Massachusetts Institute of Technology. The model's prediction results show that the mean absolute error

mean absolute percentage error

and Root Mean Squared Error are all below 1%

providing a reliable solution for the early-phase capacity prediction of in-service batteries.

关键词

Keywords

references

寇发荣, 杨天祥, 罗希, 等. 基于特征重构与多时间尺度的锂电池SOH和RUL联合估计[J]. 太阳能学报, 2025, 46(6): 68-78.

KOU F R, YANG T X, LUO X, et al. Joint estimation of SOH and RUL for lithium batteries based on feature reconstruction and multiple times scales[J]. Acta Energiae Solaris Sinica, 2025, 46(6): 68-78.

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]. 高电压技术, 2024, 50(3): 1182-1195. DOI:10.13336/j.1003-6520.hve.20221843.

XIONG Q, DI Z G, JI S C. Review on health state estimation and life prediction of lithium-ion batteries[J]. High Voltage Engineering, 2024, 50(3): 1182-1195. DOI:10.13336/j.1003-6520.hve.2022 1843.

戴俊彦, 夏明超, 陈奇芳. 基于双重注意力机制的电池SOH估计和RUL预测编解码模型[J]. 电力系统自动化, 2023, 47(6): 168-177. DOI:10.7500/AEPS20220615007.

DAI J Y, XIA M C, CHEN Q F. Encoding and decoding model of state of health estimation and remaining useful life prediction for batteries based on dual-stage attention mechanism[J]. Automation of Electric Power Systems, 2023, 47(6): 168-177. DOI:10.7500/AEPS20220615007.

程思涵, 刘思懿, 郭子旭, 等. 基于CNN-GRU-注意力的锂离子电池SOC估计[J]. 电池, 2026, 56(1): 131-135.

CHENG S H, LIU S Y, GUO Z X, et al. SOC estimation for Li-ion battery based on the CNN-GRU-Attention[J]. Battery Bimonthly, 2026, 56(1): 131-135.

蔡雨思, 李泽文, 刘萍, 等. 基于间接健康特征优化与多模型融合的锂电池SOH-RUL联合预测[J]. 电工技术学报, 2024, 39(18): 5883-5898. DOI:10.19595/j.cnki.1000-6753.tces.231057.

CAI Y S, LI Z W, LIU P, et al. Joint prediction of lithium battery state of health and remaining useful life based on indirect health features optimization and multi-model fusion[J]. Transactions of China Electrotechnical Society, 2024, 39(18): 5883-5898. DOI:10.19595/j.cnki.1000-6753.tces.231057.

黄凯, 郝润凯, 郭永芳. 基于特征综合评价和模型优化的锂离子电池健康状态估计方法[J]. 电力系统及其自动化学报, 2025, 37(5): 131-140. DOI:10.19635/j.cnki.csu-epsa.001599.

HUANG K, HAO R K, GUO Y F. State-of-health estimation method for lithium-ion battery based on comprehensive feature evaluation and model optimization[J]. Proceedings of the CSU-EPSA, 2025, 37(5): 131-140. DOI:10.19635/j.cnki.csu-epsa. 001599.

宋兴海, 张小乾, 梁惠施, 等. 基于SDAE-Transformer-ECA网络的锂电池剩余使用寿命预测[J]. 储能科学与技术, 2023, 12(10): 3181-3190. DOI:10.19799/j.cnki.2095-4239.2023.0369.

SONG X H, ZHANG X Q, LIANG H S, et al. Predicting the remaining service life of lithium batteries based on the SDAE-transformer-ECA network[J]. Energy Storage Science and Technology, 2023, 12(10): 3181-3190. DOI:10.19799/j.cnki.2095-4239.2023.0369.

SHEN S, SADOUGHI M, LI M, et al. Deep convolutional neural networks with ensemble learning and transfer learning for capacity estimation of lithium-ion batteries[J]. Applied Energy, 2020, 260: 114296. DOI:10.1016/j.apenergy.2019.114296.

YANG Z X, LI Y, YANG D, et al. Enhanced CNN-based state-of-health estimation framework for lithium-ion batteries using variable-length charging segments and transfer learning[J]. Journal of Energy Storage, 2025, 128: 117214. DOI:10.1016/j.est.2025.117214.

MA Y, SHAN C, GAO J W, et al. Multiple health indicators fusion-based health prognostic for lithium-ion battery using transfer learning and hybrid deep learning method[J]. Reliability Engineering & System Safety, 2023, 229: 108818. DOI:10.1016/j.ress.2022.108818.

GUO Y F, WANG Y S, DING P Y, et al. Future degradation trajectory prediction of lithium-ion battery based on a three-step similarity evaluation criterion for battery selection and transfer learning[J]. Journal of Energy Storage, 2023, 72: 108763. DOI:10.1016/j.est.2023.108763.

FU S Y, TAO S Y, FAN H T, et al. Data-driven capacity estimation for lithium-ion batteries with feature matching based transfer learning method[J]. Applied Energy, 2024, 353: 121991. DOI:10.1016/j.apenergy.2023.121991.

梁佳佳, 何晓霞, 肖浩逸. 基于CS-DBN的锂电池剩余寿命预测[J]. 太阳能学报, 2024, 45(3): 251-259.

LIANG J J, HE X X, XIAO H Y. Prediction of remaining useful life of lithium batteries based on CS-DBN[J]. Acta Energiae Solaris Sinica, 2024, 45(3): 251-259.

FERMÍN-CUETO P, MCTURK E, ALLERHAND M, et al. Identification and machine learning prediction of knee-point and knee-onset in capacity degradation curves of lithium-ion cells[J]. Energy and AI, 2020, 1: 100006. DOI:10.1016/j.egyai.2020. 100006.

ZHANG H, ALTAF F, WIK T. Battery capacity knee-onset identification and early prediction using degradation curvature[J]. Journal of Power Sources, 2024, 608: 234619. DOI:10.1016/j.jpowsour.2024.234619.

MARTEAU P F. Time warp edit distance with stiffness adjustment for time series matching[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(2): 306-318. DOI:10.1109/TPAMI.2008.76.

YE Z, YU J B. State-of-health estimation for lithium-ion batteries using domain adversarial transfer learning[J]. IEEE Transactions on Power Electronics, 2022, 37(3): 3528-3543. DOI:10.1109/TPEL.2021.3117788.

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. DOI:10.1038/s41560-019-0356-8.

OUYANG M S, SHEN P C. Prediction of remaining useful life of lithium batteries based on WOA-VMD and LSTM[J]. Energies, 2022, 15(23): 8918. DOI:10.3390/en15238918.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Method for monitoring irreversible lithium plating in lithium batteries using transfer learning

Automated distribution-free interval estimation for lithium-Ion battery state of health using multiple health indicators

Research progress on the industrial application of carbon nanotube conductive agents for lithium-ion batteries

State-of-health estimation for lithium-ion batteries using geometric analysis of electrochemical impedance spectroscopy

Joint frequency control strategy of distributed hydroelectric power generation and supercapacitors

Related Author

WANG Wei

LIANG Huishi

LI Miangang

ZHOU Kui

WANG Ziyao

SHI Zinan

WANG Wei

Related Institution

Tsinghua Shenzhen International Graduate School

Sichuan Energy Internet Research Institute, Tsinghua University

College of Computer Science and Technology, North University of China

College of Computer Science and Technology, Taiyuan University of Technology

Qinghai CSG New Energy Technology Co., Ltd. Delingha

⁰