计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (11A): 241200041-8.doi: 10.11896/jsjkx.241200041

• 数据库&大数据&数据科学 • 上一篇    下一篇

基于频率通道注意力机制和MSCNet的锂电池剩余使用寿命预测

卢世宇1, 王海瑞1, 朱贵富2,3, 李亚龙1   

  1. 1 昆明理工大学信息工程与自动化学院 昆明 650504
    2 昆明理工大学信息化建设管理中心 昆明 650504
    3 昆明理工大学-曙光信息产业股份有限公司AI联合研究中心 昆明 650504
  • 出版日期:2025-11-15 发布日期:2025-11-10
  • 通讯作者: 朱贵富(zhuguifu@kust.edu.cn)
  • 作者简介:lsy111315@qq.com
  • 基金资助:
    国家自然科学基金(61863016)

Remaining Useful Life Prediction of Lithium-ion Batteries Based on Frequency-channelAttention Mechanism and MSCNet

LU Shiyu1, WANG Hairui1, ZHU Guifu2,3, LI Yalong1   

  1. 1 School of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650504,China
    2 Information Construction Management Center,Kunming University of Science and Technology,Kunming 650504,China
    3 AI Joint Research Center,Kunming University of Science and Technology-Shuguang Information Industry Co.,Ltd.,Kunming 650504,China
  • Online:2025-11-15 Published:2025-11-10
  • Supported by:
    National Natural Science Foundation of China(61863016).

PDF (PC)

10

摘要： 为解决锂离子电池容量估计中特征提取不准确、数据噪声大及容量衰减趋势跟踪精度低等问题,提出了一种基于频率通道注意力机制(Frequency Channel Attention Mechanism,FCA)和MSCNet(Multi-Scale Inter-Series Correlations Net)的新型模型。模型首先对原始传感器数据进行去噪处理,以降低噪声对模型性能的干扰;其次,引入频率通道注意力机制,通过频域分析将输入序列映射到频域,识别主导时间尺度以捕捉显著的周期性模式,并对时间序列进行多尺度分解;最后,利用MSCNet对多尺度输出进行动态聚合,捕获不同时间尺度内的跨序列相关性,提升模型对时间依赖性的理解,同时减少模型参数量。在CALCE和NASA公开数据集上的实验表明,该模型在电池使用寿命预测中的相对误差(RE)较现有算法降低了10%～20%,能够更精准地跟踪电池衰退趋势。

关键词: 锂离子电池, 剩余使用寿命预测, 注意力机制, 多尺度, 跨序列相关性

Abstract: To address issues such as inaccurate feature extraction,significant data noise,and low precision in tracking capacity degradation trends in lithium-ion battery capacity estimation,a novel model combining the Frequency Channel Attention Mechanism(FCA) and Multi-Scale Inter-Series Correlations Net(MSCNet) is proposed.The model is designed in three stages.Firsty,raw sensor data are preprocessed to remove noise.Secondly,prominent periodic patterns are extracted through frequency domain analysis using the frequency-enhanced attention mechanism.Finally,the multi-scale outputs are aggregated by MSCNet,reducing model parameters while improving effectiveness of feature extraction.Experiments based on publicly available CALCE and NASA datasets demonstrate that the proposed model reduces relative error(RE) in battery life prediction by 10%~20% compared to existing algorithms,enabling more accurate tracking of battery degradation trends.

Key words: Lithiumion battery, Remaining useful life prediction, Attention mechanism, Multi-scale, Inter-series correlation

中图分类号: 

  • TM912
[1]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.
[2]LI F,MIN Y,ZHANG Y,et al.State-of-health estimation me-thod for fast-charging lithium-ion batteries based on stacking ensemble sparse Gaussian process regression[J].Reliability Engineering & System Safety,2024,242(2):1-13.
[3]WU Z Q,CHEN H J.Joint estimation of SOC and SOH of lithium battery based on adaptive H2/H∞ filtering[J].Journal of Metrology,2023,44(11):1719-1727.
[4]LI Z H,SHI Q L,WANG K L,et al.Research status and prospect of state of health estimation method for lithium-ion battery[J/OL].http://kns.cnki.net/kcms/detail/32.1180.TP.20240715.1700.006.html.
[5]NIU Z Y,JIANG X,XIE B,et al.Study on simulation and safety protection of electric vehicle overcharge and explosion accident[J].Transactions of China Electrotechnical Society,2022,37(1):36-47,57.
[6]SHU X,SHEN J W,LI G,et al.A Flexible State-of-Health Prediction Scheme for Lithium-Ion BatteryPackth Long Short-Term Memory Network and Transfer Learning[J].IEEE Transactions on Transportation Electation,2021,7(4):2238-2248.
[7]SU S S,LI W,MOU J H,et al.A Hybrid Battery Equivalent Circuit Model,Deep Learning,and Transfer Learfor Battery State Monitoring[J].IEEE Transactions on Transportation Electrification,2023,9(1):1113-1127.
[8]WU Z Q,HU X Y,MA B Y,et al.Lithium Battery SOC Estimation Based on RFF and GWO-PF [J].Journal of Quantitative & Technical Economics,2022,43(9):1200-1207.
[9]SHAHID F,ZAMEER A,MUNEEBM.A Novel Genetic LSTM Model for Wind Power Forecast[J].Energy,2021,10(1):1016-1020.
[10]DING G,WANG W,ZHU T.Remaining useful life prediction for lithium-ion batteries based on CS-VMD and GRU[J].IEEE Access,2022,10:89402-89413.
[11]XIAO S,LIU P,CHEN K,et al.Battery state of health prediction based on voltage intervals,BP neural netwnd genetic algorithm[J].International Journal of Green Energy,2024,21(8):1743-1756.
[12]JIA C,TIAN Y,SHI Y,et al.State of health prediction of lithium-ion batteries based on bidirectionalgatedurrent unit and transformer[J].Energy,2023,285:129401-129409.
[13]ZHANG M D,LIU Y,CHEN J,et al.Lithium-ion BatteryHealth State Estimation Based on ISSA-GPR [J/OL].http://kns.cnki.net/kcms/detail/12.1420.TM.20231219.1345.010.html.
[14]SONG W,WU D,SHEN W,et al.A remaining useful life prediction method for lithiumion battery based on temporal transformer network[J].Procedia Computer Science,2023,217:1830-1838.
[15]ZHOU T,MA Z,WEN Q,et al.Fedformer:Frequency enhanced decomposed transformer for long-term series forecasting[C]//International Conference on Machine Learning.PMLR,2022:27268-27286.
[16]ZHOU T,MA Z,WEN Q,et al.Film:Frequency improved legendre memory model for long-term time series forecasting[J].Advances in Neural Information Processing systems,2022,35:12677-12690.
[17]SHIZGAL B D,JUNGJ H.Towards the resolution of the Gibbs phenomena[J].Journal of Computational and Applied Mathematics,2003,161(1):41-65.
[18]SHAZEER N,MIRHOSEINI A,MAZIARZ K,et al.Outra-geously large neural networks:The sparsely-gated mixture-of-experts layer[J].arXiv:1701.06538,2017.
[19]SHEN J,MA W,SHU X,et al.Accurate state of health estimation for lithium-ion batteries under random charging scenarios[J].Energy,2023,279:128092.
[20]XIA X Y,YUE J H,ZENG X Y,et al.A Remaining Capacity Estimation Method for Lithiumion Batteries Based on State-Dependent RBF-ARX Model [J/OL].http://kns.cnki.net/kcms/detail/11.2107.TM.20240613.1329.012.html.
[21]LI C,ZHANG H L,ZHANG J P.Health Status Estimation of Spent Lithium-ion Batteries Based on Kernel Functions and Hyperparameter Optimization[J].Energy Storage Science and Technology,2024,13(6):2010-2021.
[22]WANG P,PENG X Y,CHENG Z,et al.A multi-time-scale state joint estimation method for lithiumion batteries based on data-driven model fusion [J].Automotive Engineering,2022,44(3):362-371,378.
[23]CHEN D,HONG W,ZHOU X.Transformer network for remaining useful life prediction of lithium-ion batteries[J].IEEE Access,2022,10:19621-19628.
[24]ZHOU H,LI J,ZHANG S,et al.Expanding the prediction capacity in long sequence time-series forecasting[J].Artificial Intelligence,2023,318:103886-10392.
[25]WU H,XU J,WANG J,et al.Autoformer:Decomposition transformers with auto-correlation for long-term series forecasting[J].Advances in Neural Information Processing Systems,2021,34:22419-22430.
[26]WU H,HU T,LIU Y,et al.Timesnet:Temporal 2d-variationmodeling for general time series analysis[J].arXiv:2210.02186,2023.
[27]YANG J S,FANG W GCHEN J Y,et al.A lithium-ion battery remaining useful life prediction method based on unscented particle filter and optimal combination strategy[J].Journal of Energy Storage,2022,55:105648.
[28]ZHANG Y,CHEN L,LI Y,et al.A hybrid approach for remaining useful life prediction of lithiumion battery with adaptive levy flight optimized particle filter and long shortterm memory network[J].Journal of Energy Storage,2021,44:103245.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发