基于早期老化与迁移学习的电池寿命预测研究

doi:10.3969/j.issn.1008-0198.2025.01.001

湖南电力 ›› 2025, Vol. 45 ›› Issue (1): 1-7.doi: 10.3969/j.issn.1008-0198.2025.01.001

• 专家专栏：电化学储能系统关键技术与应用 • 上一篇下一篇

基于早期老化与迁移学习的电池寿命预测研究

杨天, 沈进冉, 官亦标, 高飞, 姜添, 刘庆

中国电力科学研究院有限公司,北京 100192

收稿日期:2024-11-01 修回日期:2024-11-26 出版日期:2025-02-25 发布日期:2025-03-05
通信作者: 官亦标(1980),男,正高级工程师,主要从事新型储能技术研究工作。
作者简介:杨天(1997),男,工程师,从事新型储能试验评价研究工作。沈进冉（1989）,女,工程师,从事新型储能试验评价研究工作。高飞（1981）,男,正高级工程师,从事电池储能试验评价研究工作。姜添(2001),男,硕士研究生,主要从事储能电池试验评价研究工作。刘庆(1991),男,工程师,从事新型储能试验评价研究工作。
基金资助:
国家电网有限公司科技项目（5419-202323785A-3-8-KJ）

Research on Battery Life Prediction Based on Early Aging and Transfer Learning

YANG Tian, SHEN Jinran, GUAN Yibiao, GAO Fei, JIANG Tian, LIU Qing

China Electric Power Research Institute, Beijing 100192, China

Received:2024-11-01 Revised:2024-11-26 Online:2025-02-25 Published:2025-03-05

摘要/Abstract

摘要： 针对储能电池寿命预测可靠性不足、泛化性不佳的问题,提出一种基于LSTM的储能电池寿命预测方法。预测模型仅采用电池前150个循环的早期老化数据作为输入,通过滑动窗口法训练模型,实现了储能电池寿命的精准预测,1 000个循环内均方根误差为0.11%。同时开发迁移学习模块,对于新型号储能电池,采用第150次循环至250次循环对基础模型进行迁移修正,将电池寿命预测均方根误差由0.50%降至0.10%,提升了寿命预测模型的泛化性,为寿命预测技术在储能电池生产及检测评价领域的实际应用提供一种解决方法。

关键词: 储能电池, 寿命预测, 迁移学习, 早期老化数据

Abstract: To address the problems of insufficient reliability and poor generalization in life prediction of energy storage batteries, a large-capacity energy storage battery remaining life prediction method based on LSTM is proposed, which only adopts the early aging data of the batteryfrom the first 150 cycles as the inputand realizes accurate prediction of the remaining life of energy storage batteries through the model training method of the sliding-window method. This approach achieves a remarkable root mean square error (RMSE) of 0.11% over a full national standard testing period of 1000 cycles. Additionally, a transfer learning module is developedand for the new model of energy storage batteries, the 150th to 250th cycles are used for the migration correction of the basic model, which reduces RMSE of the battery life prediction from 0.50% to 0.10%. This advancement enhances the generalization of the life prediction model, providing a solution for the practical application of the life prediction technology in the field of energy storage battery production, testing and evaluation.

Key words: energy storage battery, life prediction, transfer learning, early aging data

中图分类号:

TM911

杨天, 沈进冉, 官亦标, 高飞, 姜添, 刘庆. 基于早期老化与迁移学习的电池寿命预测研究[J]. 湖南电力, 2025, 45(1): 1-7.

YANG Tian, SHEN Jinran, GUAN Yibiao, GAO Fei, JIANG Tian, LIU Qing. Research on Battery Life Prediction Based on Early Aging and Transfer Learning[J]. Hunan Electric Power, 2025, 45(1): 1-7.

参考文献

[1] 官亦标,沈进冉,刘家亮,等. 以安全高质量应用为导向的储能锂离子电池综合性能评价标准[J]. 储能科学与技术,2023,12(9):2946-2953.
[2] 李相俊,赵珊珊,惠东. 面向新型电力系统的大型储能电站关键技术发展趋势分析与展望[J]. 供用电,20,39(7):2-8.
[3] HU Y G,LIANG J D,CHEN X X,et al.Comparative study of thermodynamic & kinetic parameters measuring techniques in lithium-ion batteries[J]. Journal of Power Sources,2024,606:234590.
[4] TIAN J P,XIONG R,SHEN W X,et al.Data-driven battery degradation prediction:Forecasting voltage-capacity curves using one-cycle data[J]. EcoMat,2022,4(5):e12213.
[5] CHEN C,XIONG R,YANG R X,et al.A novel data-driven method for mining battery open-circuit voltage characterization[J]. Green Energy and Intelligent Transportation,2022,1(1):100001.
[6] SULZER V,MOHTAT P,AITIO A,et al.The challenge and opportunity of battery lifetime prediction from field data[J]. Joule,2021,5(8):1934-1955.
[7] TIAN J P,XIONG R,LU J H,et al.Battery state-of-charge estimation amid dynamic usage with physics-informed deep learning[J]. Energy Storage Materials,2022(50):718-729.
[8] 柳应全,王岩冰,郭赟. 高倍率脉冲放电锂电池的电气参数辨识[J]. 海军工程大学学报,2024,36(5):41-47.
[9] 巫春玲,赵玉冰,马耀,等. 采用改进最大相关熵自适应迭代容积卡尔曼滤波算法的锂离子电池荷电状态估计[J]. 西安交通大学学报,2024,58(11):52-64.
[10] 朱振威,苗嘉伟,祝夏雨,等. 基于机器学习方法的锂电池剩余寿命预测研究进展[J]. 储能科学与技术,2024,13(9):3134-349
[11] 耿萌萌,范茂松,杨凯,等. 基于EIS和神经网络的退役电池SOH快速估计[J]. 储能科学与技术,2022,11(2):673-678.
[12] TIAN J P,XIONG R,SHEN W X,et al.Deep neural network battery charging curve prediction using 30 points collected in 10 min[J]. Joule,2021,5(6):1521-1534.
[13] WANG J Y,ZHANG C P,ZHANG L J,et al.A novel aging characteristics-based feature engineering for battery state of health estimation[J]. Energy,2023,273:127169.
[14] 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):383-91.
[15] RAN A H,CHENG M,CHEN S X,et al.Fast remaining capacity estimation for lithium-ion batteries based on short-time pulse test and Gaussian process regression[J]. Energy & Environmental Materials,2023,6(3):e12386.
[16] ZHAO J Y,LING H P,LIU J,et al.Machine learning for predicting battery capacity for electric vehicles[J]. eTransportation,2023,15:100214.
[17] 国家市场监督管理总局,国家标准化管理委员会. 电力储能用锂离子电池:GB/T 36276—2023[S]. 北京:中国标准出版社,2023.
[18] 贾新羽. 锂离子电池老化模式诊断与寿命快速评估方法 [D]. 北京:北京交通大学,2023.
[19] 李卓昊,石琼林,王康丽,等. 锂离子电池健康状态估计方法研究现状与展望[J]. 电力系统自动化,2024,48(20):109-129.
[20] 田成文,孙丙香,赵鑫泽,等. 基于数据驱动的锂离子电池快速寿命预测[J]. 储能科学与技术,2024,13(9):3103-3111.
[21] 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.
[22] MA G J,XU S P,JIANG B B,et al.Real-time personalized health status prediction of lithium-ion batteries using deep transfer learning[J]. Energy & Environmental Science,2022,15(10):4083-4094.

基于早期老化与迁移学习的电池寿命预测研究

Research on Battery Life Prediction Based on Early Aging and Transfer Learning

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 1

编辑推荐

Metrics

本文评价