容量再生干扰下锂离子电池剩余使用寿命有监督离线回放预测方法

廖志伟; 孔博; 吕东泽

doi:10.19799/j.cnki.2095-4239.2026.0657

您当前的位置：

首页 >

文章列表页 >

容量再生干扰下锂离子电池剩余使用寿命有监督离线回放预测方法

储能科学与技术 | 更新时间：2026-10-16

- 容量再生干扰下锂离子电池剩余使用寿命有监督离线回放预测方法
- Supervised Offline Replay Prediction Method for Remaining Useful Life of Lithium-Ion Batteries under Capacity Regeneration Interference
- 储能科学与技术 2026年页码：1-17
- 作者机构：
  
  华南理工大学电力学院，广东广州 510630
- 作者简介：
  
  廖志伟（1973—），男，博士，副教授，锂离子电池寿命预测与健康管理，epliao@scut.edu.cn。
- 基金信息：
  
  广东省基础与应用基础研究基金项目(2023A1515240038)
- DOI：10.19799/j.cnki.2095-4239.2026.0657
  中图分类号： TM 912
- 收稿：2026-07-27，
  
  修回：2026-10-15，
  
  网络首发：2026-10-16，
- 稿件说明：
移动端阅览
廖志伟, 孔博, 吕东泽. 容量再生干扰下锂离子电池剩余使用寿命有监督离线回放预测方法[J]. 储能科学与技术, XXXX, XX(XX): 1-17.

LIAO Zhiwei, KONG Bo, LYU Dongze. Supervised Offline Replay Prediction Method for Remaining Useful Life of Lithium-Ion Batteries under Capacity Regeneration Interference[J]. Energy Storage Science and Technology, XXXX, XX(XX): 1-17.
廖志伟, 孔博, 吕东泽. 容量再生干扰下锂离子电池剩余使用寿命有监督离线回放预测方法[J]. 储能科学与技术, XXXX, XX(XX): 1-17. DOI： 10.19799/j.cnki.2095-4239.2026.0657.

LIAO Zhiwei, KONG Bo, LYU Dongze. Supervised Offline Replay Prediction Method for Remaining Useful Life of Lithium-Ion Batteries under Capacity Regeneration Interference[J]. Energy Storage Science and Technology, XXXX, XX(XX): 1-17. DOI： 10.19799/j.cnki.2095-4239.2026.0657.

摘要

针对锂离子电池容量再生引起的局部非单调变化、跨电池退化差异及目标域有效窗口不足，提出一种有监督离线回放剩余使用寿命（remaining useful life，RUL）预测方法。在完整寿命记录条件下，目标电池容量观测按原始循环顺序释放，并遵循“先预测、后评价、再更新”的时序。基于容量差分的指数加权移动平均（exponentially weighted moving average，EWMA）机制用于标记容量再生候选区间、筛选连续有效窗口并维持区间内RUL输出连续；门控循环单元（gated recurrent unit，GRU）经源域预训练后，在目标域冻结第一层并更新第二层及预测层；质量门控变分自编码器生成对抗网络（variational autoencoder generative adversarial network，VAEGAN）作为探索性增强分支。NASA电池数据集上的多随机种子、消融、敏感性和增强对照实验表明，分层适配可降低部分电池的回放误差，EWMA参数影响有效窗口和最终误差，VAEGAN未显示相对于简单增强的稳定增量收益，GRU在当前配置下兼顾误差水平与参数规模。所提框架可用于分析容量再生条件下不同处理与适配环节的时序作用。

Abstract

A supervised offline replay method is proposed for remaining useful life (RUL) prediction under capacity-regeneration-induced local non-monotonicity

cross-battery degradation discrepancies

and limited valid target-domain windows. When complete-life records are available

target-battery observations are released in chronological order following a predict-evaluate-update sequence. An exponentially weighted moving average (EWMA) mechanism based on capacity differences identifies candidate regeneration intervals

screens continuous valid windows

and maintains RUL continuity within these intervals. A gated recurrent unit (GRU) model is pretrained in the source domain; during target-domain adaptation

its first layer is frozen while the second layer

fully connected layer

and output layer are updated. A quality-gated variational autoencoder generative adversarial network (VAEGAN) is included as an exploratory augmentation branch. Multi-seed

ablation

sensitivity

and augmentation-comparison experiments on NASA batteries show that layer-wise adaptation reduces replay errors for some batteries

EWMA parameters affect valid-window availability and final errors

VAEGAN provides no stable incremental gain over simple augmentation

and GRU offers a favorable error-parameter balance under the current configuration. The framework supports reproducible analysis of the temporal effects of processing and adaptation strategies under capacity regeneration.

关键词

Keywords

references

LI T, ZHOU Z, THELEN A, et al. Predicting battery lifetime under varying usage conditions from early aging data[J]. Cell Reports Physical Science, 2024, 5(4): 101891. DOI: 10.1016/j.xcrp.2024.101891.

ZHANG Y, WIK T, BERGSTRÖM J, et al. A machine learning-based framework for online prediction of battery ageing trajectory and lifetime using histogram data[J]. Journal of Power Sources, 2022, 526: 231110. DOI: 10.1016/j.jpowsour.2022.231110.

ATTIA P M, BILLS A, BROSA PLANELLA F, et al. Review—'Knees' in lithium-ion battery aging trajectories[J]. Journal of The Electrochemical Society, 2022, 169(6): 060517. DOI: 10.1149/1945-7111/ac6d13.

吴胜利, 郭琪, 邢文婷. 考虑容量再生的可解释性锂离子电池剩余寿命预测方法[J]. 电工技术学报, 2026, 41(9): 3208-3220. DOI: 10.19595/j.cnki.1000-6753.tces.250665.

HE N, YANG Z, QIAN C, et al. Remaining useful life prediction of lithium-ion battery based on fusion model considering capacity regeneration phenomenon[J]. Journal of Energy Storage, 2024, 85: 111068. DOI: 10.1016/j.est.2024.111068.

MA Q, ZHENG Y, YANG W, et al. Remaining useful life prediction of lithium battery based on capacity regeneration point detection[J]. Energy, 2021, 234: 121233. DOI: 10.1016/j.energy.2021.121233.

CHEN X, LIU Z, SHENG H, et al. Transfer learning based remaining useful life prediction of lithium-ion battery considering capacity regeneration phenomenon[J]. Journal of Energy Storage, 2024, 76: 109798. DOI: 10.1016/j.est.2023.109798.

LI Z, SHEN S, YE Y, et al. An interpretable online prediction method for remaining useful life of lithium-ion batteries[J]. Scientific Reports, 2024, 14: 12541. DOI: 10.1038/s41598-024-63160-2.

WANG X H, DAI K, HU M, et al. Lithium-ion battery health state and remaining useful life prediction based on hybrid model MFE-GRU-TCA[J]. Journal of Energy Storage, 2024, 95: 112442. DOI: 10.1016/j.est.2024.112442.

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.

董作林, 宋金岩, 孟子迪. 基于模态分解和深度学习的锂离子电池寿命预测[J]. 储能科学与技术, 2025, 14(4): 1645-1653. DOI: 10.19799/j.cnki.2095-4239.2024.0983.

陈勇, 王俊磊, 王鹏, 等. 锂离子电池早期剩余寿命预测方法综述[J]. 电池, 2026, 56(1): 222-230. DOI: 10.19535/j.1001-1579.2026.01.035.

CHE Y, STROE D I, HU X, et al. Semi-supervised self-learning-based lifetime prediction for batteries[J]. IEEE Transactions on Industrial Informatics, 2023, 19(5): 6471-6481. DOI: 10.1109/TII.2022.3206776.

MA L, TIAN J, ZHANG T, et al. Enhanced battery life prediction with reduced data demand via semi-supervised representation learning[J]. Journal of Energy Chemistry, 2025, 101: 524-534. DOI: 10.1016/j.jechem.2024.10.001.

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.

CHOWDHURY M A, MODEKWE G, LU Q. Lithium-ion battery capacity prediction via conditional recurrent generative adversarial network-based time-series regeneration[C]//Proceedings of the 2025 American Control Conference. IEEE, 2025: 1469-1475. DOI: 10.23919/ACC63710.2025.11107734.

SOO Y Y, WANG Y, XIANG H, et al. A data augmentation method for lithium-ion battery capacity estimation based on Wasserstein Time Generative Adversarial Network[J]. Energy Technology, 2025, 13(8): 2400488. DOI: 10.1002/ente.202400488.

LIU K, PENG Q, CHE Y, et al. Transfer learning for battery smarter state estimation and ageing prognostics: Recent progress, challenges, and prospects[J]. Advances in Applied Energy, 2023, 9: 100117. DOI: 10.1016/j.adapen.2022.100117.

LEWERENZ M, DECHENT P, SAUER D U. Investigation of capacity recovery during rest period at different states-of-charge after cycle life test for prismatic Li(Ni1/3Mn1/3Co1/3)O2-graphite cells[J]. Journal of Energy Storage, 2019, 21: 680-690. DOI: 10.1016/j.est.2019.01.004.

EPDING B, RUMBERG B, JAHNKE H, et al. Investigation of significant capacity recovery effects due to long rest periods during high current cyclic aging tests in automotive lithium ion cells and their influence on lifetime[J]. Journal of Energy Storage, 2019, 22: 249-256. DOI: 10.1016/j.est.2019.02.015.

LIU Z, WANG Y, CHEN J, et al. Dual-Driven Battery Life Prediction Based on Data Augmentation and Capacity Recovery Identification[J]. Particuology, 2026. DOI: 10.1016/j.partic.2026.08.006.

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

深度学习框架在锂离子电池充电状态估计的应用与优化

多特征融合的MVMD-VSN_GRU锂离子电池健康状态预测方法研究

基于集成学习和分位数回归的锂离子电池寿命预测方法

改进Sine混沌映射CO-ELM锂离子电池RUL预测

基于EEMD-GRU-NN锂离子电池表面温度预测方法研究