基于多通道特征融合优化VMD-SCNN-LSTM的锂电池RUL预测

勇晔; 吴参林; 薛溟枫; 毛晓波

doi:doi:10.19912/j.0254-0096.tynxb.2024-1457

English Version Reader LoginJournal Login

您当前的位置：

首页 >

文章列表页 >

基于多通道特征融合优化VMD-SCNN-LSTM的锂电池RUL预测

更新时间：2026-01-08

- 基于多通道特征融合优化VMD-SCNN-LSTM的锂电池RUL预测
- Vol. 46, Issue 12, Pages: 41-51(2025)
- 作者机构：
- 作者简介：
- 基金信息：
- DOI：doi:10.19912/j.0254-0096.tynxb.2024-1457
  CLC： TM912
- Published Online：07 January 2026，
  
  Published：2025
- 稿件说明：
移动端阅览
勇晔, 吴参林, 薛溟枫, et al. 基于多通道特征融合优化VMD-SCNN-LSTM的锂电池RUL预测[J]. 2025, 46(12): 41-51.
DOI：

勇晔, 吴参林, 薛溟枫, et al. 基于多通道特征融合优化VMD-SCNN-LSTM的锂电池RUL预测[J]. 2025, 46(12): 41-51. DOI： doi:10.19912/j.0254-0096.tynxb.2024-1457.

摘要

针对电池容量非线性衰退问题

提出一种基于多通道特征融合优化VMD-SCNN-LSTM的锂电池剩余使用寿命（RUL）预测方法。首先

根据皮尔逊相关系数选取健康因子（HI）

提取锂电池充放电过程中电压、电流和温度的时序特性曲线。其次

采用变分模态分解（VMD）对电池容量劣化曲线进行分解。最后

将HIs和经过VMD分解的容量频域特征作为卷积神经网络（CNN）的多通道并行输入

原始容量作为长短期记忆网络（LSTM）的输入

再将两者提取的特征进行融合开展RUL预测。同时引入3倍交叉验证训练方法和dropout技术来避免过拟合问题。实验结果表明

所提最优混合模型的均方根误差RMSE不超过3.1%

平均绝对百分比误差MAPE不超过1.3%。

Abstract

Aiming at the problem of nonlinear decline of battery capacity

a RUL prediction method for lithium battery based on multi-channel feature fusion optimization VMD-SCNN-LSTM is proposed. Firstly

the health factors are selected according to the Pearson correlation coefficient

and the time-series characteristic curves of voltage

current and temperature during the charging and discharging process of Li-ion battery are extracted. Secondly

the battery capacity degradation curves are decomposed using variational modal decomposition （VMD）. Finally

the HI and the frequency-domain features of the capacity decomposed by VMD are used as the multi-channel parallel inputs to the convolutional neural network （CNN）

and the original capacity was used as the input to the long-short-term memory network （LSTM）

and then the two extracted features are fused to carry out RUL prediction. A triple cross-validation training method and dropout technique are also introduced to avoid the overfitting problem. The optimal hybrid model proposed in this study has an RMSE of no more than 3.1% and a MAPE of no more than 1.3%.

关键词

Keywords

references

赵珈卉, 田立亭, 程林. 锂离子电池状态估计与剩余寿命预测方法综述[J]. 发电技术, 2023, 44(1): 1-17.

夏向阳, 吕崇耿, 吴小忠, 等. 基于混合模型的储能用锂离子电池剩余寿命预测方法研究[J]. 太阳能学报, 2024, 45(10): 726-735.

吴诗淼, 王文波, 朱婷, 等. 基于跳跃连接多尺度CNN的锂离子电池剩余寿命预测[J]. 太阳能学报, 2024, 45(7): 199-208.

LI X Y, CHEN L, HUA W, et al.Optimal charging for lithium-ion batteries to avoid lithium plating based on ultrasound-assisted diagnosis and model predictive control[J]. Applied energy, 2024, 367: 123396.

CHEN C Q, HUANG Y F, YU X Y, et al.Improving the accuracy of voltage estimation in the low charge state range at low temperature: an equivalent circuit model considering the influence of temperature on solid phase diffusion process[J]. Journal of energy storage, 2024, 88: 111577.

MERROUCHE W, LEKOUAGHET B, BOUGUENNA E, et al.Parameter estimation of ECM model for Li-ion battery using the weighted mean of vectors algorithm[J]. Journal of energy storage, 2024, 76: 109891.

VIGNESH R, ASHOK B.Intelligent energy management through neuro-fuzzy based adaptive ECMS approach for an optimal battery utilization in plugin parallel hybrid electric vehicle[J]. Energy conversion and management, 2023, 280: 116792.

MOHAMMADI F.Lithium-ion battery state-of-charge estimation based on an improved Coulomb-counting algorithm and uncertainty evaluation[J]. Journal of energy storage, 2022, 48: 104061.

CHE Y H, HU X S, LIN X K, et al.Health prognostics for lithium-ion batteries: mechanisms, methods, and prospects[J]. Energy & environmental science, 2023, 16(2): 338-371.

ZHENG X J, FANG H J.An integrated unscented Kalman filter and relevance vector regression approach for lithium-ion battery remaining useful life and short-term capacity prediction[J]. Reliability engineering & system safety, 2015, 144: 74-82.

WANG L Z, JIANG S Y, MAO Y T, et al.Lithium-ion battery state of health estimation method based on variational quantum algorithm optimized stacking strategy[J]. Energy reports, 2024, 11: 2877-2891.

SHE C Q, LI Y, ZOU C F, et al.Offline and online blended machine learning for lithium-ion battery health state estimation[J]. IEEE transactions on transportation electrification, 2022, 8(2): 1604-1618.

何冰琛, 杨薛明, 王劲松, 等. 基于PCA-GPR的锂离子电池剩余使用寿命预测[J]. 太阳能学报, 2022, 43(5): 484-491.

LI Z H, BAI F, ZUO H F, et al.Remaining useful life prediction for lithium-ion batteries based on iterative transfer learning and mogrifier LSTM[J]. Batteries, 2023, 9(9): 448.

QI X, HONG C F, YE T, et al.Frequency reconstruction oriented EMD-LSTM-AM based surface temperature prediction for lithium-ion battery[J]. Journal of energy storage, 2024, 84: 111001.

ZHANG Z, LIU X, DONG X, et al.A hybrid RUL prediction framework for lithium-ion batteries based on EEMD and KAN-LSTM[J]. Batteries, 2025, 11(10): 348.

LI C Z, LIN W, WU H Y, et al.Performance degradation decomposition-ensemble prediction of PEMFC using CEEMDAN and dual data-driven model[J]. Renewable energy, 2023, 215: 118913.

YUAN Z F, TIAN T, HAO F C, et al.A hybrid neural network based on variational mode decomposition denoising for predicting state-of-health of lithium-ion batteries[J]. Journal of power sources, 2024, 609: 234697.

DING G R, WANG W B, ZHU T.Remaining useful life prediction for lithium-ion batteries based on CS-VMD and GRU[J]. IEEE access, 2022, 10: 89402-89413.

KATTENBORN T, LEITLOFF J, SCHIEFER F, et al.Review on convolutional neural networks (CNN) in vegetation remote sensing[J]. ISPRS journal of photogrammetry and remote sensing, 2021, 173: 24-49.

HAFIZHAHULLAH H, YULIANI A R, PARDEDE H, et al.A hybrid CNN-LSTM for battery remaining useful life prediction with charging profiles data[C]//Proceedings of the 2022 International Conference on Computer, Control, Informatics and Its Applications. Virtual Event, Indonesia, 2023: 106-110.

CHEN J C, CHEN T L, LIU W J, et al.Combining empirical mode decomposition and deep recurrent neural networks for predictive maintenance of lithium-ion battery[J]. Advanced engineering informatics, 2021, 50: 101405.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

基于GSS的RIME-CNN-LSTM-Attention光伏出力组合预测模型

基于VMD和多时间尺度分类预测的风电单机短期功率预测研究

RESEARCH ON SO-CNN-LSTM PHOTOVOLTAIC POWER PREDICTION MODEL BASED ON COMPREHENSIVE SIMILAR DAY SELECTION

PHOTOVOLTAIC POWER PREDICTION METHOD COMBINATING SOLAR RADIATION CALCULATION AND CNN-LSTM

ADAPTIVE ULTRA-SHORT-TERM WIND SPEED PREDICTION MODEL CONSIDERING ERROR INFORMATION

Related Author

黄毅

马一力

苗安康

袁越

夏卫平

邓艾东

薛原

卞文彬

Related Institution

College of Environmental Science and Engineering, North China Electric Power University

Chinese Academy of Environmental Planning

Department of Automation, North China Electric Power University

School of Economics and Management, North China Electric Power University

Economic and Technology Research Institute, State Grid Chongqing Electric Power Company

AI问答

⁰