基于故障特征增强与混合神经网络的锂电池内短路状态估算研究

doi:10.19595/j.cnki.1000-6753.tces.251084

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (0 KB) HTML
输出: BibTeX | EndNote (RIS)

摘要及时检测并精准识别电池组内发生内短路的故障电池,对于保障电动汽车及电池储能系统的安全稳定运行具有重要意义。该文提出了一种基于故障特征增强与混合神经网络的锂离子电池组内短路状态估计方法,在保证高精度的同时有效地提升了算法效率。首先,面向内短路早期电压差异微弱的特点,构建平均-差异模型,通过计算每个单体电池与“平均电池”之间的端电压差,构建差异电压特征序列,该特征能突出单体相对于电池组整体行为的异常偏移,有效反映微小的电池状态变化,从而显著提高内短路征兆在特征空间的可检测性,实现针对性的故障特征增强。在此基础上,基于该模型在仿真平台中对不同内短路程度进行系统性建模与参数注入,生成三种故障级别的高可信模拟故障数据集,并与实验故障数据相融合,有效地扩展了故障特征样本,为后续状态分类奠定了坚实的基础。随后,在使用卷积神经网络（CNN）提取局部空间特征基础上,引入双向长短时记忆网络（BiLSTM）以捕捉时序信息,构建混合神经网络模型实现对内短路故障特征的深层建模,并对不同程度的内短路状态进行识别。最后,开展了锂电池内短路模拟实验,从模型精度、不同模型性能对比以及数据增强的有效性三个方面对所提方法进行了试验验证。结果表明,所构建的模型在微短路、中度短路和严重短路工况下的分类准确率分别达到96.0%、98.3%和99.0%,充分验证了该算法在内短路故障识别任务中的准确性与有效性。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	申江卫
	徐攀
	舒星
	魏福星
	陈峥

关键词 ：锂离子电池, 状态估计, 混合神经网络, 数据增强, 内短路

Abstract：Timely detection and accurate identification of faulted batteries with internal short circuits are crucial for ensuring the safe and reliable operation of electric vehicles (EVs) and energy storage systems (ESS). Early detection of such faults helps to prevent safety hazards such as thermal runaway or fire, thereby ensuring the safety of both the battery system and the user. This paper proposes a novel method for estimating internal short-circuit states in lithium-ion battery packs employing a hybrid neural network approach. The proposed method effectively improves the algorithm's efficiency while maintaining high classification accuracy.
The method first addresses the subtle voltage differences that occur in the early stages of internal short circuits, where the voltage drop is often minimal. An average-difference model is constructed to compute the voltage difference between each battery cell and an “average cell”, yielding a sequence of differential voltage features. This sequence highlights the abnormal behavior of a single cell relative to the battery pack's overall behavior. It can detect even the slightest deviations in battery behavior, which would otherwise be difficult to distinguish from normal operation.
Then, different internal short-circuit severities are simulated. By injecting various parameters, high- confidence simulated fault datasets are generated to represent three distinct fault levels: micro, moderate, and severe short-circuit stages. These simulated datasets are then combined with experimental fault data, yielding a larger, more diverse sample of fault features. This enhanced dataset allows the model to learn from a richer set of fault features, thereby improving its ability to generalize across different battery conditions.
To further improve fault detection, a hybrid neural network model is incorporated using convolutional neural networks (CNNs) to extract local spatial features within the battery pack and bidirectional long short-term memory (BiLSTM) networks to capture temporal information from time-series data. The BiLSTM network processes a sequence of voltage differences over time, enabling the model to identify patterns and trends in fault progression. This hybrid approach leverages both spatial and temporal features, enabling the model to classify internal short-circuit states more accurately.
The proposed method is validated through internal short-circuit simulation experiments. The performance is evaluated from three perspectives: model accuracy, comparisons among different models, and the effectiveness of data augmentation techniques. The results show that the hybrid model achieves classification accuracies of 96.0%, 98.3%, and 99.0% for the micro-, moderate-, and severe-short-circuit stages, respectively. The proposed method is highly effective at detecting and identifying internal short-circuit faults in lithium-ion battery packs.
In conclusion, the proposed method offers a robust and efficient approach to detecting internal short-circuit faults in lithium-ion battery systems. By combining fault feature enhancement with a hybrid CNN-BiLSTM neural network, the method improves fault classification accuracy. It provides an effective solution for safety- critical applications in electric vehicles and energy storage systems. This approach not only enables early detection of internal short-circuit faults but also contributes to the development of safer, more reliable battery management systems.

Key words： Lithium-ion battery state estimation hybrid neural network data augmentation internal short circuit

收稿日期: 2025-06-19

PACS:

TM912

基金资助:国家自然科学基金（52577236）、云南省基础研究计划（202301AT070423）、云南省重点研发计划“云南省智能网联交通国际联合实验室”（202503AP140016）和重庆市自然科学基金（CSTB2024NSCQ-MSX0389）资助项目

通讯作者: 申江卫男,1984年生,教授,硕士生导师,研究方向为新能源汽车动力电池管理。E-mail: shenjiangwei6@kust.edu.cn

作者简介: 陈峥男,1982年生,教授,博士生导师,研究方向为新能源汽车节能控制与动力电池管理、智能网联汽车优化控制。E-mail: chen@kust.edu.cn

引用本文:

申江卫, 徐攀, 舒星, 魏福星, 陈峥. 基于故障特征增强与混合神经网络的锂电池内短路状态估算研究[J]. 电工技术学报, 2026, 41(12): 4298-4312. Shen Jiangwei, Xu Pan, Shu Xing, Wei Fuxing, Chen Zheng. Internal Short Circuit State Estimation for Lithium-Ion Batteries Based on Fault Feature Augmentation and Hybrid Neural Network Modeling. Transactions of China Electrotechnical Society, 2026, 41(12): 4298-4312.

链接本文:

https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.251084 https://dgjsxb.ces-transaction.com/CN/Y2026/V41/I12/4298

[1] 安富强, 赵洪量, 程志, 等. 纯电动车用锂离子电池发展现状与研究进展[J]. 工程科学学报, 2019, 41(1): 22-42.
An Fuqiang, Zhao Hongliang, Cheng Zhi, et al.Development status and research progress of power battery for pure electric vehicles[J]. Chinese Journal of Engineering, 2019, 41(1): 22-42.
[2] Kong Xiangdong, Lu Languang, Yuan Yuebo, et al.Foreign matter defect battery and sudden spontaneous combustion[J]. eTransportation, 2022, 12: 100170.
[3] 牛志远, 姜欣, 谢镔, 等. 电动汽车过充燃爆事故模拟及安全防护研究[J]. 电工技术学报, 2022, 37(1): 36-47, 57.
Niu Zhiyuan, Jiang Xin, Xie Bin, 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.
[4] Xu Chaojie, Li Laibao, Xu Yuwen, et al.A vehicle-cloud collaborative method for multi-type fault diagnosis of lithium-ion batteries[J]. eTransportation, 2022, 12: 100172.
[5] 耿莉敏, 赵扬, 高楠, 等. 锂离子电池内短路形成机理及检测方法综述[J]. 汽车工程学报, 2023, 13(4): 481-495.
Geng Limin, Zhao Yang, Gao Nan, et al.Review of short-circuit formation mechanisms and diagnostic methods in lithium-ion batteries[J]. Chinese Journal of Automotive Engineering, 2023, 13(4): 481-495.
[6] Feng X, He X, Lu L, et al.Analysis on the fault features for internal short circuit detection using an electrochemical-thermal coupled model[J]. Journal of the Electrochemical Society, 2018, 165(2): A155-A167.
[7] Gao Wenkai, Zheng Yuejiu, Ouyang Minggao, et al.Micro-short-circuit diagnosis for series-connected lithium-ion battery packs using mean-difference model[J]. IEEE Transactions on Industrial Electronics, 2019, 66(3): 2132-2142.
[8] Wang Lu, Ma Jian, Zhao Xuan, et al.Online estimation of state-of-charge inconsistency for lithium-ion battery based on SVSF-VBL[J]. Journal of Energy Storage, 2023, 67: 107657.
[9] 赵帅. 锂离子动力电池内短路机理与诊断方法研究[D]. 长春: 吉林大学, 2022.
Zhao Shuai.Study on mechanism and diagnosis method of internal short circuit in lithium-ion power battery[D]. Changchun: Jilin University, 2022.
[10] 谢国民, 刘澳. 基于混合特征提取和机器学习的锂离子电池健康状态估计[J]. 电力系统自动化, 2025, 49(21): 120-130.
Xie Guomin, Liu Ao.State of health estimation for lithium-ion batteries based on hybrid feature extraction and machine learning[J]. Automation of Electric Power Systems, 2025, 49(21): 120-130.
[11] Naha A, Khandelwal A, Agarwal S, et al.Internal short circuit detection in Li-ion batteries using supervised machine learning[J]. Scientific Reports, 2020, 10: 1301.
[12] Xia B, Nguyen T, Yang J, et al.The improved interleaved voltage measurement method for series connected battery packs[J]. Journal of Power Sources, 2016, 334: 12-22.
[13] 王君瑞, 吴新举, 赵东琦, 等. 基于WOA-VMD与PSO-SVM的锂离子电池内短路故障诊断方法[J]. 工程科学学报, 2023, 45(12): 2162-2172.
Wang Junrui, Wu Xinju, Zhao Dongqi, et al.Research on internal short-circuit fault diagnosis methods for lithium-ion batteries based on WOA-VMD and PSO-SVM[J]. Chinese Journal of Engineering, 2023, 45(12): 2162-2172.
[14] Shang Yunlong, Lu Gaopeng, Kang Yongzhe, et al.A multi-fault diagnosis method based on modified Sample Entropy for lithium-ion battery strings[J]. Journal of Power Sources, 2020, 446: 227275.
[15] Wang Zhenpo, Hong Jichao, Liu Peng, et al.Voltage fault diagnosis and prognosis of battery systems based on entropy and Z-score for electric vehicles[J]. Applied Energy, 2017, 196: 289-302.
[16] 彭自然, 王顺豪, 肖伸平, 等. 基于KA Informer的电动汽车动力电池荷电状态和健康状态估算[J]. 电工技术学报, 2025, 40(19): 6378-6394.
Peng Ziran, Wang Shunhao, Xiao Shenping, et al.State of charge and state of health estimation of electric vehicle power battery based on KA Informer model[J]. Transactions of China Electrotechnical Society, 2025, 40(19): 6378-6394.
[17] 郭煜, 王亦伟, 彭鹏, 等. 基于孤立森林算法的锂离子电池微内短路故障诊断方法[J]. 储能科学与技术, 2024, 13(11): 4102-4112.
Guo Yu, Wang Yiwei, Peng Peng, et al.Fault diagnosis of micro-internal short circuits in lithium-ion battery using the isolated forest algorithm[J]. Energy Storage Science and Technology, 2024, 13(11): 4102-4112.
[18] Yang Sen, Xu Boran, Peng Hanlin.Isolation and grading of faults in battery packs based on machine learning methods[J]. Electronics, 2022, 11(9): 1494.
[19] Sun Tao, Zhu Hao, Xu Yuwen, et al.Internal short circuit fault diagnosis for the lithium-ion batteries with unknown parameters based on transfer learning optimized residual network by multi-label data processing[J]. Journal of Cleaner Production, 2024, 444: 141224.
[20] 卫志农, 原康康, 成乐祥, 等. 基于多新息最小二乘算法的锂电池参数辨识[J]. 电力系统自动化, 2019, 43(15): 139-145.
Wei Zhinong, Yuan Kangkang, Cheng Lexiang, et al.Parameter identification of lithium-ion battery based on multi-innovation least squares algorithm[J]. Auto-mation of Electric Power Systems, 2019, 43(15): 139-145.
[21] Hu Xiaosong, Li Shengbo, Peng Huei, et al.Robu-stness analysis of state-of-charge estimation methods for two types of Li-ion batteries[J]. Journal of Power Sources, 2012, 217: 209-219.
[22] 向兆军, 胡凤玲, 罗明华, 等. 基于电池组模型和聚类算法的锂离子电池组SOC不一致估计[J]. 机械工程学报, 2020, 56(18): 154-163.
Xiang Zhaojun, Hu Fengling, Luo Minghua, et al.Estimation of SOC inconsistencies in lithium-ion battery packs based on battery pack modeling and clustering algorithm[J]. Journal of Mechanical Engineering, 2020, 56(18): 154-163.
[23] 高文凯. 锂离子动力电池的短路故障诊断研究[D]. 上海: 上海理工大学, 2020.
Gao Wenkai.Study on short circuit diagnosis of lithium ion power battery[D]. Shanghai: University of Shanghai for Science & Technology, 2020.
[24] 张芳芳, 强科亮, 娄本杰. 基于US06复合工况的动力电池SOC估计[J]. 电工技术, 2022(18): 69-71.
Zhang Fangfang, Qiang Keliang, Lou Benjie.SOC estimation of power battery based on US06 composite condition[J]. Electric Engineering, 2022(18): 69-71.
[25] 刘力硕, 张明轩, 卢兰光, 等. 锂离子电池内短路机理与检测研究进展[J]. 储能科学与技术, 2018, 7(6): 1003-1015.
Liu Lishuo, Zhang Mingxuan, Lu Languang, et al.Recent progress on mechanism and detection of internal short circuit in lithium-ion batteries[J]. Energy Storage Science and Technology, 2018, 7(6): 1003-1015.
[26] 范兴明, 封浩, 张鑫. 最小二乘算法优化及其在锂离子电池参数辨识中的应用[J]. 电工技术学报, 2024, 39(5): 1577-1588.
Fan Xingming, Feng Hao, Zhang Xin.Optimization of least squares method and its application in parameter identification of lithium-ion battery model[J]. Transac-tions of China Electrotechnical Society, 2024, 39(5): 1577-1588.
[27] 杜伟, 王圣, 李健, 等. 基于CNN-LSTM-AM模型的储能锂离子电池荷电状态预测[J]. 电工技术学报, 2025, 40(9): 2982-2995.
Du Wei, Wang Sheng, Li Jian, et al.Prediction of state of charge for energy storage lithium-ion batteries based on CNN-LSTM-AM model[J]. Transactions of China Electrotechnical Society, 2025, 40(9): 2982-2995.
[28] 黄凯, 丁恒, 郭永芳, 等. 基于数据预处理和长短期记忆神经网络的锂离子电池寿命预测[J]. 电工技术学报, 2022, 37(15): 3753-3766.
Huang Kai, Ding Heng, Guo Yongfang, 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.
[29] 张开放, 苏华友, 窦勇. 一种基于混淆矩阵的多分类任务准确率评估新方法[J]. 计算机工程与科学, 2021, 43(11): 1910-1919.
Zhang Kaifang, Su Huayou, Dou Yong.A new multi-classification task accuracy evaluation method based on confusion matrix[J]. Computer Engineering and Science, 2021, 43(11): 1910-1919.