高阶矩匹配的动态分布自适应电池组故障诊断方法

官思伟; 何志伟; 董哲康; 童宏涛; 马沈辉; 高明裕

doi:10.11999/JEIT250226

高阶矩匹配的动态分布自适应电池组故障诊断方法

doi: 10.11999/JEIT250226 cstr: 32379.14.JEIT250226

官思伟¹,
何志伟^{1, 2, ,},
董哲康^{1, 2},
童宏涛¹,
马沈辉¹,
高明裕^{1, 2}

1.
杭州电子科技大学电子信息学院杭州 310018
2.
全省智能汽车电子研究重点实验室杭州 310018

基金项目: 国家重点研发计划(2023YFB2504100)，国家自然科学基金(62171170, 62227802)

详细信息

作者简介:
官思伟：博士生，研究方向为锂电池故障诊断、健康状态估计

何志伟：教授，研究方向为新能源汽车、电池管理系统

董哲康：副教授，研究方向为电池管理、深度学习

童宏涛：博士生，研究方向为锂电池故障诊断、残值估计

马沈辉：博士生，研究方向为锂电池异常检测

高明裕：教授，研究方向为电池管理系统、能源转换与储能

通讯作者:
何志伟　zwhe@hdu.edu.cn

中图分类号: TN911; TP183
计量
- 文章访问数: 17
- HTML全文浏览量: 13
- PDF下载量: 2
- 被引次数: 0
出版历程
- 收稿日期: 2025-04-01
- 修回日期: 2025-09-12
- 网络出版日期: 2025-09-16

Dynamic Distribution Adaptation with Higher-Order Moment Matching for Battery Pack Multi-Fault Diagnosis

GUAN Siwei¹,
HE Zhiwei^{1, 2
, ,},
DONG Zhekang^{1, 2},
TONG Hongtao¹,
MA Shenhui¹,
GAO Mingyu^{1, 2}

1.
School of Electronics Information, Hangzhou Dianzi University, Hangzhou 310018, China
2.
Provincial Key Laboratory of Intelligent Automotive Electronics Research, Hangzhou 310018, China

Funds: The National Key Research and Development Program of China (2023YFB2504100), The National Natural Science Foundation of China (62171170, 62227802)

摘要

摘要: 锂离子电池作为是一种具有高度复杂电化学反应的系统，为了满足电动汽车的功率和能量需求，需要将大量电池串并联组成电池组。然而，电池组的安全问题成为其广泛应用的关键挑战。现有电池组诊断方法在实际应用中存在不足，主要受限于运行条件的多变性和故障样本的稀缺性。此外，电池组电压呈现复杂的非高斯分布，使得基于差异的领域自适应方法仅能表征有限的故障统计特征。针对上述问题，该文提出一种高阶矩匹配的动态分布自适应电动汽车电池组多故障诊断方法，该方法在源域中学习可迁移特征，从而实现目标域中的故障诊断，可诊断的故障类型包括内部短路、电压传感器漂移故障、电压传感器噪声故障和电池不一致故障。所提方法通过动态因子评估边缘和条件分布的相对重要性，动态学习域不变特征。此外，该方法还利用高阶矩匹配对非高斯分布的电池放电特征进行精细化领域对齐。在3种不同的电动汽车标准运行工况下进行的跨域故障诊断实验结果表明，该方法优于现有的基线方法，并实现了平均 95% F1分数的故障诊断性能。
- 电池故障诊断 /
- 新能源汽车 /
- 动态分布自适应 /
- 高阶矩匹配
Abstract: Objective Electric vehicle battery pack fault diagnosis is challenged by diverse operating conditions, the scarcity of fault data, and the domain shift caused by the non-Gaussian distribution of battery features. Conventional fault diagnosis methods struggle to address multiple fault types, lack the capability for fault isolation, and fail to account for distribution shifts between training and test data. Domain adaptation approaches enable robust multi-fault diagnosis across operating conditions without relying on accurate cell models or abundant labeled data. However, current methods remain limited. (1) They typically assume that aligning global and fine-grained subdomain distributions is equally important, which may not hold in practice. (2) Knowledge transfer cannot be fully achieved by aligning only low-order statistical features; higher-order statistical features are needed to capture the non-Gaussian characteristics of battery discharge profiles. To address these issues, a method is proposed in which global domains and subdomains are dynamically aligned while higher-order statistical moments are extracted to represent complex non-Gaussian distributions, thereby achieving fine-grained domain alignment and effective knowledge transfer. Methods This study proposes a dynamic distribution adaptation method with higher-order moment matching for multi-fault diagnosis of battery packs. The approach consists of three components: (1) Dynamic distribution adaptation. A feature extractor based on a one-dimensional convolutional network with residual connectivity and a multilayer perceptron classifier is constructed. The global distributions of source and target domains are aligned using Maximum Mean Discrepancy (MMD), while subdomain distributions of similar faults are aligned using Local Maximum Mean Discrepancy (LMMD). A dynamic factor is introduced to automatically adjust the relative weights of global and local alignment according to the inter-domain discrepancy, thereby adapting to distribution shifts under different operating conditions. (2) Higher-order moment matching. To address the non-Gaussian characteristics of battery data, higher-order statistical moment matching is incorporated into MMD. Computational complexity in high-dimensional tensors is reduced by random sampling, which enables fine-grained domain alignment across multi-order statistics and enhances the transferability of non-Gaussian distribution features. (3) Multi-fault diagnosis with domain adaptation. Experimental data from three standard vehicle operating conditions are used to jointly optimize classification loss and domain adaptation loss. This enables the diagnosis of multiple faults, including internal short circuit, sensor drift/noise, and battery inconsistency, across operating conditions while reducing reliance on manual annotation. By dynamically integrating global and local feature alignment, the method improves generalization performance under complex operating conditions and non-Gaussian distribution scenarios. Results and Discussions Systematic experiments validate the superiority of the proposed dynamic distribution adaptation with higher-order moment matching for multi-fault diagnosis in electric vehicle battery packs. As shown in Table 3, the results from six transfer tasks under three operating conditions demonstrate that the proposed method achieves an average F1 score of 95%, which is 13.3% higher than that of the best-performing baseline model (DSAN). The confusion matrix in Fig. 6 indicates that the method achieves the lowest misclassification rate in distinguishing similar faults. Feature visualization results (Fig. 7) show that the method effectively reduces cross-domain feature distances of similar faults by dynamically adjusting the weights of global and local distribution alignment. Moreover, it successfully captures non-Gaussian discharge characteristics through higher-order moment matching, thereby achieving fine-grained domain adaptation. In terms of efficiency, the proposed method attains an average diagnosis time of 0.404 3 seconds (Table 4), satisfying real-time on-board application requirements. Nonetheless, optimization of computational resource consumption remains necessary for deployment on edge devices. Importantly, the method does not require labeled data from the target domain and overcomes the generalization bottleneck of traditional methods under domain shift and non-Gaussian conditions. However, some cross-domain features (Fig. 7) are not completely overlapped, and lightweight model design is still required for practical implementation on edge devices. Conclusions The battery pack is recognized as a critical component of electric vehicles, and reliable multi-fault diagnosis is regarded as essential for safe operation. Considering the unknown and diverse nature of real operating conditions, fault diagnosis is investigated across three driving cycles: UDDS, FUDS, and US06. A dynamic distribution adaptation with higher-order moment matching (DDAMD) is proposed for diagnosing multiple faults in series-connected battery packs. The method dynamically evaluates the relative importance of conditional and marginal distributions to align source and target domains, while non-Gaussian features from charge–discharge curves are effectively extracted for fine-grained alignment. Experimental results across six transfer tasks confirm that DDAMD achieves the highest diagnostic performance. Detailed analyses present diagnostic accuracy for each fault type as well as the diagnostic speed, while feature visualization further improves interpretability by demonstrating how the algorithm extracts domain-invariant and discriminative fault features across domains. Future research will extend this work in two directions: (1) incorporating additional operating conditions and a broader set of fault categories, and (2) exploring transfer tasks from simulation to real-world applications to facilitate data acquisition and labeling.
- Battery fault diagnosis /
- Electric vehicles /
- Dynamic distribution adaptation /
- Higher-order moment matching

HTML全文

图 1 DDAMD电池故障诊断架构。

下载: 全尺寸图片幻灯片

图 2 实验平台

下载: 全尺寸图片幻灯片

图 3 3种汽车标准运行工况

下载: 全尺寸图片幻灯片

图 4 在FUDS运行工况下的电池包电压曲线

下载: 全尺寸图片幻灯片

图 5 所有模型在6个迁移任务的准确率(%)

下载: 全尺寸图片幻灯片

图 6 模型在FUDS→US06迁移任务的混淆矩阵

下载: 全尺寸图片幻灯片

图 7 所有模型在US06→FUDS迁移任务源域和目标域样本的特征分布

下载: 全尺寸图片幻灯片

表 1 电池规格

参数	配置
尺寸	18.6×65.2 mm
额定电压	3.7 V
额定容量	2 000 mAh
充电截止电压	4.2 V
放电截止电压	2.5 V
重量	48 g
内阻	≤60 mΩ

下载: 导出CSV

表 2 故障参数具体设置

类型	描述	参数
内短路	电池内部短路	1/5/10 Ω
传感器漂移故障	传感器受到低频信号干扰	1 Hz
传感器噪声故障	电压传感器受噪声干扰	0.1 V
不一致故障	在电池组中容量不一致	0.2 V

下载: 导出CSV

表 3 模型在6个迁移任务上的性能对比结果(%)

方法	UDDS→FUDS			UDDS→US06			FUDS→US06
方法	精度	召回率	F1分数	精度	召回率	F1分数	精度	召回率	F1分数
WDCNN	79.00	77.37	77.75	78.92	75.85	74.96	90.28	89.99	89.92
DANN	76.40	77.09	76.57	78.89	78.28	78.46	90.98	87.70	86.81
CDAN	73.67	72.44	70.99	77.33	76.66	74.06	86.95	84.61	83.38
DSAN	87.22	87.38	87.21	84.01	83.13	82.16	95.56	94.93	94.91
CRDAN	79.47	80.53	79.76	79.42	79.78	79.40	87.13	86.08	85.53
DDAMD	93.09	92.35	92.25	95.62	95.53	95.51	98.81	98.71	98.72
方法	UDDS→FUDS			UDDS→US06			FUDS→US06
方法	精度	召回率	F1分数	精度	召回率	F1分数	精度	召回率	F1分数
WDCNN	79.65	69.68	67.06	68.60	63.60	64.25	93.20	92.31	92.28
DANN	68.61	68.02	65.36	84.19	77.56	76.43	94.80	94.58	94.57
CDAN	86.46	80.93	77.17	89.26	87.37	87.39	94.79	94.46	94.47
DSAN	85.48	75.78	74.73	79.14	60.47	54.43	96.09	96.05	96.04
CRDAN	87.10	77.56	75.18	79.17	77.84	74.55	90.34	88.17	87.89
DDAMD	90.43	88.54	87.96	94.74	94.65	94.66	99.89	99.88	99.88

下载: 导出CSV

表 4 所有模型的平均诊断时间(s)

模型	WDCNN	DANN	CDAN	DSAN	CRDAN	DDAMD
诊断时间	0.192 3	0.187 3	0.201 6	0.414 2	0.203 2	0.404 3

下载: 导出CSV

参考文献(35)

[1]	XU Yiming, GE Xiaohua, GUO Ruohan, et al. Recent advances in model-based fault diagnosis for lithium-ion batteries: A comprehensive review[J]. Renewable and Sustainable Energy Reviews, 2025, 207: 114922. doi: 10.1016/j.rser.2024.114922.
[2]	DONG Zhekang, GU Shenyu, ZHOU Shiqi, et al. Periodic segmentation transformer-based internal short circuit detection method for battery packs[J]. IEEE Transactions on Transportation Electrification, 2025, 11(1): 3655–3666. doi: 10.1109/TTE.2024.3444453.
[3]	YU Quanqing, YANG Yu, TANG Aihua, et al. Unsupervised learning for lithium-ion batteries fault diagnosis and thermal runaway early warning in real-world electric vehicles[J]. Journal of Energy Storage, 2025, 109: 115194. doi: 10.1016/j.est.2024.115194.
[4]	张照娓, 郭天滋, 高明裕, 等. 电动汽车锂离子电池荷电状态估算方法研究综述[J]. 电子与信息学报, 2021, 43(7): 1803–1815. doi: 10.11999/JEIT200487. ZHANG Z haowei, GUO Tianzi, GAO Mingyu, et al. Review of SoC estimation methods for electric vehicle Li-ion batteries[J]. Journal of Electronics & Information Technology, 2021, 43(7): 1803–1815. doi: 10.11999/JEIT200487.
[5]	GUAN Siwei, HE Zhiwei, MA Shenhui, et al. Early life prediction of fast-charging battery based on feature engineering and state space models[J]. Journal of Energy Storage, 2025, 127: 116969. doi: 10.1016/j.est.2025.116969.
[6]	高明裕, 蔡林辉, 孙长城, 等. 一种基于斯皮尔曼秩相关结合神经网络的电池组内部短路故障检测算法[J]. 电子与信息学报, 2022, 44(11): 3734–3747. doi: 10.11999/JEIT210975. GAO Mingyu, CAI Linhui, SUN Changcheng, et al. An internal short circuit fault detecting of battery pack based on spearman rank correlation combined with neural network[J]. Journal of Electronics & Information Technology, 2022, 44(11): 3734–3747. doi: 10.11999/JEIT210975.
[7]	HONG Zhongshen, WANG Yujie, and JIN Zhichao. Diagnosis of battery external short circuits based on an improved second-order RC fault model and recursive least squares identification method[J]. Energy, 2025, 319: 134880. doi: 10.1016/j.energy.2025.134880.
[8]	CHENG Zhixiang, MIN Yuanyuan, QIN Peng, et al. A distributed thermal-pressure coupling model of large-format lithium iron phosphate battery thermal runaway[J]. Applied Energy, 2025, 378: 124875. doi: 10.1016/j.apenergy.2024.124875.
[9]	XIONG Rui, SUN Xinjie, MENG Xiangfeng, et al. Advancing fault diagnosis in next-generation smart battery with multidimensional sensors[J]. Applied Energy, 2024, 364: 123202. doi: 10.1016/j.apenergy.2024.123202.
[10]	ZHAO Rui, LIU Jie, and GU Junjie. Simulation and experimental study on lithium ion battery short circuit[J]. Applied Energy, 2016, 173: 29–39. doi: 10.1016/j.apenergy.2016.04.016.
[11]	XIONG Rui, YU Quanqing, SHEN Weixiang, et al. A sensor fault diagnosis method for a lithium-ion battery pack in electric vehicles[J]. IEEE Transactions on Power Electronics, 2019, 34(10): 9709–9718. doi: 10.1109/TPEL.2019.2893622.
[12]	LIU Zhentong and HE Hongwen. Sensor fault detection and isolation for a lithium-ion battery pack in electric vehicles using adaptive extended Kalman filter[J]. Applied Energy, 2017, 185: 2033–2044. doi: 10.1016/j.apenergy.2015.10.168.
[13]	XIA Bing, SHANG Yunlong, NGUYEN T, et al. A correlation based fault detection method for short circuits in battery packs[J]. Journal of Power Sources, 2017, 337: 1–10. doi: 10.1016/j.jpowsour.2016.11.007.
[14]	WANG Shunli, TANG Wu, FERNANDEZ C, et al. A novel endurance prediction method of series connected lithium-ion batteries based on the voltage change rate and iterative calculation[J]. Journal of Cleaner Production, 2019, 210: 43–54. doi: 10.1016/j.jclepro.2018.10.349.
[15]	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. doi: 10.1016/j.jpowsour.2019.227275.
[16]	ZHANG Zhendong, KONG Xiangdong, ZHENG Yuejiu, et al. Real-time diagnosis of micro-short circuit for Li-ion batteries utilizing low-pass filters[J]. Energy, 2019, 166: 1013–1024. doi: 10.1016/j.energy.2018.10.160.
[17]	WANG Ping, CHEN Jiqing, LAN Fengchong, et al. Multiscale feature fusion approach to early fault diagnosis in EV power battery using operational data[J]. Journal of Energy Storage, 2024, 98: 112812. doi: 10.1016/j.est.2024.112812.
[18]	YANG Ruixin, XIONG Rui, HE Hongwen, et al. A fractional-order model-based battery external short circuit fault diagnosis approach for all-climate electric vehicles application[J]. Journal of Cleaner Production, 2018, 187: 950–959. doi: 10.1016/j.jclepro.2018.03.259.
[19]	ZHAO Yang, LIU Peng, WANG Zhenpo, et al. Fault and defect diagnosis of battery for electric vehicles based on big data analysis methods[J]. Applied Energy, 2017, 207: 354–362. doi: 10.1016/j.apenergy.2017.05.139.
[20]	ZHANG Wei, PENG Gaoliang, LI Chuanhao, et al. A new deep learning model for fault diagnosis with good anti-noise and domain adaptation ability on raw vibration signals[J]. Sensors, 2017, 17(2): 425. doi: 10.3390/s17020425.
[21]	QIU Xianghui, BAI Yu, and WANG Shuangfeng. A novel unsupervised domain adaptation-based method for lithium-ion batteries state of health prognostic[J]. Journal of Energy Storage, 2024, 75: 109358. doi: 10.1016/j.est.2023.109358.
[22]	GUAN Siwei, HE Zhiwei, MA Shenhui, et al. Domain adaptation with contrastive learning for lithium-ion battery packs fault diagnosis[J]. IEEE Transactions on Transportation Electrification, 2025. doi: 10.1109/TTE.2025.3582411. (查阅网上资料,未找到本条文献卷期页码信息,请确认).
[23]	CAI Linhui, WANG Han, DONG Zhekang, et al. A multi-fault diagnostic method based on category-reinforced domain adaptation network for series-connected battery packs[J]. Journal of Energy Storage, 2023, 60: 106690. doi: 10.1016/j.est.2023.106690.
[24]	ZHAO Zhibin, ZHANG Qiyang, YU Xiaolei, et al. Applications of unsupervised deep transfer learning to intelligent fault diagnosis: A survey and comparative study[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 3525828. doi: 10.1109/TIM.2021.3116309.
[25]	ZHU Yongchun, ZHUANG Fuzhen, WANG Jindong, et al. Deep subdomain adaptation network for image classification[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 32(4): 1713–1722. doi: 10.1109/TNNLS.2020.2988928.
[26]	ZHANG Jing, LI Wanqing, and OGUNBONA P. Joint geometrical and statistical alignment for visual domain adaptation[C]. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, USA, 2017: 5150–5158. doi: 10.1109/CVPR.2017.547.
[27]	WANG Jindong, CHEN Yiqiang, FENG Wenjie, et al. Transfer learning with dynamic distribution adaptation[J]. ACM Transactions on Intelligent Systems and Technology (TIST), 2020, 11(1): 6. doi: 10.1145/3360309.
[28]	XIANG Xiaowei, LIU Yang, FANG Gaoyun, et al. Two-stage alignments framework for unsupervised domain adaptation on time series data[J]. IEEE Signal Processing Letters, 2023, 30: 698–702. doi: 10.1109/LSP.2023.3264621.
[29]	CHEN Chao, FU Zhihang, CHEN Zhihong, et al. HoMM: Higher-order moment matching for unsupervised domain adaptation[C]. Proceedings of the 34th AAAI Conference on Artificial Intelligence, New York, USA, 2020: 3422–3429. doi: 10.1609/aaai.v34i04.5745.
[30]	KANG Yongzhe, YANG Xichen, ZHOU Zhongkai, et al. A comparative study of fault diagnostic methods for lithium-ion batteries based on a standardized fault feature comparison method[J]. Journal of Cleaner Production, 2021, 278: 123424. doi: 10.1016/j.jclepro.2020.123424.
[31]	SUN Jinlei, CHEN Siwen, XING Shiyou, et al. A battery internal short circuit fault diagnosis method based on incremental capacity curves[J]. Journal of Power Sources, 2024, 602: 234381. doi: 10.1016/j.jpowsour.2024.234381.
[32]	KANG Yongzhe, DUAN Bin, ZHOU Zhongkai, et al. A multi-fault diagnostic method based on an interleaved voltage measurement topology for series connected battery packs[J]. Journal of Power Sources, 2019, 417: 132–144. doi: 10.1016/j.jpowsour.2019.01.058.
[33]	LAI Xin, YI Wei, KONG Xiangdong, et al. Online detection of early stage internal short circuits in series-connected lithium-ion battery packs based on state-of-charge correlation[J]. Journal of Energy Storage, 2020, 30: 101514. doi: 10.1016/j.est.2020.101514.
[34]	GANIN Y and LEMPITSKY V. Unsupervised domain adaptation by backpropagation[C]. Proceedings of the 32nd International Conference on International Conference on Machine Learning, Lille, France, 2015: 1180–1189.
[35]	LONG Mingsheng, CAO Zhangjie, WANG Jianmin, et al. Conditional adversarial domain adaptation[C]. Proceedings of the 32nd International Conference on Neural Information Processing Systems, Montréal, Canada, 2018: 1647–1657.