电源学报

基于迁移学习模型的IGBT老化参数预测方法

PDF下载

罗韡 , 刘宇凡 , 张德稳

电源学报 | 人工智能与电力电子 2026,24(6): 301-310

收起

电源学报 |人工智能与电力电子 2026 , 24 (6) : 301 -310

基于迁移学习模型的IGBT老化参数预测方法

全屏

[Author(id=1322531791131497215, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=luowei@usst.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1322531791261520641, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, authorId=1322531791131497215, language=EN, stringName=Wei LUO, firstName=Wei, middleName=null, lastName=LUO, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322531791391544066, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, authorId=1322531791131497215, language=CN, stringName=罗韡, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=上海理工大学机械工程学院，上海　200093, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322531790963725051, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, xref=null, ext=[AuthorCompanyExt(id=1322531790984696572, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, companyId=1322531790963725051, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China), AuthorCompanyExt(id=1322531791014056701, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, companyId=1322531790963725051, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=上海理工大学机械工程学院，上海　200093)])]), Author(id=1322531791508984580, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=lyfsmailbox@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322531793157346054, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, authorId=1322531791508984580, language=EN, stringName=Yufan LIU, firstName=Yufan, middleName=null, lastName=LIU, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322531793220260615, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, authorId=1322531791508984580, language=CN, stringName=刘宇凡, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=上海理工大学机械工程学院，上海　200093, bio={"content":"

刘宇凡（2000—　），男，硕士研究生。研究方向：微弱信号检测及功率器件可靠性。E-mail：lyfsmailbox@163.com。

"}, bioImg=null, bioContent=

刘宇凡（2000—　），男，硕士研究生。研究方向：微弱信号检测及功率器件可靠性。E-mail：lyfsmailbox@163.com。

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1322531790963725051, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, xref=null, ext=[AuthorCompanyExt(id=1322531790984696572, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, companyId=1322531790963725051, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China), AuthorCompanyExt(id=1322531791014056701, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, companyId=1322531790963725051, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=上海理工大学机械工程学院，上海　200093)])]), Author(id=1322531793291563785, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=dw_zhang21@163.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1322531793367061259, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, authorId=1322531793291563785, language=EN, stringName=Dewen ZHANG, firstName=Dewen, middleName=null, lastName=ZHANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1322531793446753036, tenantId=1146029695717560320, journalId=1146031654075715584, articleId=1322531789638324984, authorId=1322531793291563785, language=CN, stringName=张德稳, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=上海理工大学机械工程学院，上海　200093, bio={"content":"

张德稳（1998—　），男，硕士研究生。研究方向：功率器件可靠性。E-mail：dw_zhang21@163.com。

"}, bioImg=null, bioContent=

张德稳（1998—　），男，硕士研究生。研究方向：功率器件可靠性。E-mail：dw_zhang21@163.com。

罗韡, 刘宇凡, 张德稳

作者信息

上海理工大学机械工程学院，上海　200093

通讯作者:

罗韡（1986—　），男，通信作者，博士，副教授。研究方向：多电平变流器，大功率同步电机控制，物联网。E-mail：luowei@usst.edu.cn。

作者简介:

刘宇凡（2000—　），男，硕士研究生。研究方向：微弱信号检测及功率器件可靠性。E-mail：lyfsmailbox@163.com。

张德稳（1998—　），男，硕士研究生。研究方向：功率器件可靠性。E-mail：dw_zhang21@163.com。

Prediction Method for IGBT Aging Parameter Based on Transfer Learning Model

Wei LUO, Yufan LIU, Dewen ZHANG

Affiliations

School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China

出版时间: 2026-06-30 doi: 10.13234/j.issn.2095-2805.2026.6.301

文章导航

摘要

收起

针对绝缘栅双极型晶体管IGBT（insulated gate bipolar transistor）的健康状态及剩余使用寿命RUL（remaining useful life）评估问题，采用基于迁移学习的自适应循环神经网络AdaRNN（adaptive recurrent neural network）模型预测IGBT的老化参数。首先分析IGBT的失效机理，选择集电极-发射极导通电压作为评价IGBT健康状态的老化参数；然后搭建IGBT老化实验台，在线采集该老化参数；最后基于采集的老化参数，利用AdaRNN模型实现IGBT老化参数预测。结果表明，使用同一IGBT老化参数作为训练集与测试集时，传统时序模型和AdaRNN模型预测结果的精度与速度相似；使用2组不同IGBT的老化参数作为训练集与测试集时，AdaRNN模型可解决由于器件差异导致的老化参数差异问题，预测的准确度明显高于传统时序模型，模型的泛化能力和鲁棒性强。

关键词

绝缘栅双极型晶体管 / 数据驱动 / 老化参数 / 迁移学习

Abstract

收起

Aimed at the health status of insulated gate bipolar transistors (IGBTs) and the problem of evaluation on their remaining useful life (RUL), the aging parameter of IGBTs was predicted based on an adaptive recurrent neural network (AdaRNN) transfer learning model. First, the failure mechanism of IGBTs was analyzed, and the on-state collector-emitter voltage was selected as the aging parameter to evaluate the health status of IGBTs. Then, an IGBT aging test bench was set up to collect the aging parameter online. Finally, based on the acquired aging parameter, the AdaRNN model was used to predict the IGBT aging parameter. Results show that when the same IGBT aging parameter was used as the training and test sets, the accuracy and speed of prediction results of the traditional time series model and AdaRNN model were similar. When the aging parameters of two sets of different IGBTs were used as the training and test sets, the AdaRNN model can solve the problem of aging parameter difference caused by device difference, and the prediction accuracy was obviously higher than that of the traditional time series model, indicating the strong generalization capability and robustness of the new model.

Key words

Insulated gate bipolar transistor (IGBT) / data-driven / aging parameter / transfer learning

引用本文

罗韡, 刘宇凡, 张德稳. 基于迁移学习模型的IGBT老化参数预测方法. 电源学报, 2026 , 24 (6) : 301 -310 . DOI: 10.13234/j.issn.2095-2805.2026.6.301

Wei LUO, Yufan LIU, Dewen ZHANG. Prediction Method for IGBT Aging Parameter Based on Transfer Learning Model[J]. Journal of Power Supply, 2026 , 24 (6) : 301 -310 . DOI: 10.13234/j.issn.2095-2805.2026.6.301

参考文献

收起

文献

收起

参考文献引证文献

排序方式：

[1]

Alghassi A, Perinpanayagam S, Samie M. Stochastic RUL calculation enhanced with TDNN-based IGBT failure modeling[J]. IEEE Transactions on Reliability, 2016, 65(2): 558-573.

[2]

赵蕤, 杨柯欣, 唐涛, 等. 基于一种新型导通压降采样电路的IGBT器件结温在线监测方法[J]. 电源学报, 2024, 22(6): 295-303. Zhao Rui, Yang Kexin, Tang Tao, et al. Online monitoring method for IGBT device junction temperature based on novel on-state voltage drop sampling circuit[J]. Journal of Power Supply, 2024, 22(6): 295-303. (in Chinese)

[3]

Batard C, Ginot N, Antonios J. Lumped dynamic electrothermal model of IGBT module of inverters[J]. IEEE Transactions on Components, Packaging and Manufacturing Technology, 2015, 5(3): 355-364.

[4]

Tang Yunyu, Ma Hao. Dynamic electrothermal model of paralleled IGBT modules with unbalanced stray parameters[J]. IEEE Transactions on Power Electronics, 2017, 32(2): 1385-1399.

[5]

Deng Erping, Zhang Jingwei, Zhao Zhibin, et al. Influence of the clamping force on the power cycling lifetime reliability of press pack IGBT sub-module[J]. The Journal of Engineering, 2019, 2019(16): 2435-2439.

[6]

敬敏, 程禹, 陈学锋. 基于电热耦合模型的功率器件热可靠性仿真分析[J]. 现代雷达, 2022, 44(9): 93-98. Jing Min, Cheng Yu, Chen Xuefeng. Simulation and analysis of power device’s thermal reliability based on electro-thermal modeling[J]. Modern Radar, 2022, 44(9): 93-98. (in Chinese)

[7]

白梁军, 黄萌, 饶臻, 等. 基于GARCH模型的IGBT寿命预测[J]. 中国电机工程学报, 2020, 40(18): 5787-5795. Bai Liangjun, Huang Meng, Rao Zhen, et al. Lifetime prediction of IGBT based on GARCH model[J]. Proceedings of the CSEE, 2020, 40(18): 5787-5795. (in Chinese)

[8]

Ali S H, Heydarzadeh M, Dusmez S, et al. Lifetime estimation of discrete IGBT devices based on Gaussian process[J]. IEEE Transactions on Industry Applications, 2018, 54(1): 395-403.

[9]

裴洪, 胡昌华, 司小胜, 等. 基于机器学习的设备剩余寿命预测方法综述[J]. 机械工程学报, 2019, 55(8): 1-13. Pei Hong, Hu Changhua, Si Xiaosheng, et al. Review of machine learning based remaining useful life prediction methods for equipment[J]. Journal of Mechanical Engineering, 2019, 55(8): 1-13. (in Chinese)

[10]

王飞, 黄涛, 杨晔. 基于Stacking多模型融合的IGBT器件寿命的机器学习预测算法研究[J]. 计算机科学, 2022, 49(S1): 784-789. Wang Fei, Huang Tao, Yang Ye. Study on machine learning algorithms for life prediction of IGBT devices based on stacking multi-model fusion[J]. Computer Science, 2022, 49(S1): 784-789. (in Chinese)

[11]

任宏宇, 余瑶怡, 杜雄, 等. 基于优化长短期记忆神经网络的IGBT寿命预测模型[J]. 电工技术学报, 2024, 39(4): 1074-1086. Ren Hongyu, Yu Yaoyi, Du Xiong, et al. IGBT lifetime prediction model based on optimized long short-term memory neural network[J]. Transactions of China Electrotechnical Society, 2024, 39(4): 1074-1086. (in Chinese)

[12]

朱琴跃, 于逸尘, 占岩文, 等. 基于短时傅里叶变换和深度网络的模块化多电平换流器子模块IGBT开路故障诊断[J]. 电工技术学报, 2024, 39(12): 3840-3854. Zhu Qinyue, Yu Yichen, Zhan Yanwen, et al. IGBT open-circuit fault diagnosis of modular multilevel converter sub-module based on short-time Fourier transform and deep networks[J]. Transactions of China Electrotechnical Society, 2024, 39(12): 3840-3854. (in Chinese)

[13]

杨强. 迁移学习[M]. 北京: 机械工业出版社, 2020. Yang Qiang. Transfer Learning[M]. Beijing: China Machine Press, 2020. (in Chinese)

[14]

Salinas D, Flunkert V, Gasthaus J, et al. DeepAR: Probabilistic forecasting with autoregressive recurrent networks[J]. International Journal of Forecasting, 2020, 36(3): 1181-1191.

[15]

Le Guen V, Thome N. Shape and time distortion loss for training deep time series forecasting models[C]//33rd Conference on Neural Information Processing Systems (NeurIPS 2019). Vancouver, Canada, 2019: 1-13.

[16]

肖凯, 王振, 严喜林, 等. 压接式IGBT健康管理方法综述[J]. 电源学报, 2024, 22(3): 199-210. Xiao Kai, Wang Zhen, Yan Xilin, et al. Review of health management methods for press-pack IGBTs[J]. Journal of Power Supply, 2024, 22(3): 199-210. (in Chinese)

[17]

石巍. 大功率IGBT模块健康状态信息提取方法研究及加速老化试验平台研制[D]. 杭州: 浙江大学, 2018. Shi Wei. Research on the extraction method of health status information for high-power IGBT modules and development of an accelerated aging test platform. Hangzhou: Zhejiang University, 2018. (in Chinese)

[18]

Zhang Yihui, Liu Yinghan, Li Cui, et al. Analysis of fault precursor parameters under accelerated aging tests for IGBT modules[C]//2020 17th China International Forum on Solid State Lighting & 2020 International Forum on Wide Bandgap Semiconductors China (SSLChina: IFWS). Shenzhen, China, 2021: 238-241.

2026年第24卷第6期

PDF下载

引用本文

BibTeX

文章信息

doi: 10.13234/j.issn.2095-2805.2026.6.301

接收时间：2024-03-03
首发时间：2026-10-29
出版时间：2026-06-30

补充材料

相关文章

文章信息

作者

出版历史

收稿日期：2024-03-03
修回日期：2024-06-22
录用日期：2024-08-02

基金

作者信息

上海理工大学机械工程学院，上海　200093

通讯作者:

罗韡（1986—　），男，通信作者，博士，副教授。研究方向：多电平变流器，大功率同步电机控制，物联网。E-mail：luowei@usst.edu.cn。

参考文献

分享链接

https://castjournals.cast.org.cn/joweb/dyxb/CN/10.13234/j.issn.2095-2805.2026.6.301

分享至

全文二维码

扫描看全文

引用本文

BibTeX

本文的引用情况

2种不同金属材料的力学参数

科 Family	属数 Number of genus	种数 Number of species	占总种数比例 Percentage of total species (%)	属 Genus	种数 Number of species	占总种数比例 Percentage of total species (%)
鹅膏菌科Amanitaceae	2	11	5.26	鹅膏菌属 Amanita	10	4.78
小菇科 Mycenaceae	2	12	5.74	丝盖伞属 Inocybe	5	2.39
多孔菌科 Polyporaceae	8	14	6.70	蜡蘑属 Laccaria	5	2.39
红菇科 Russulaceae	3	23	11.00	小皮伞属 Marasmius	6	2.87
				小菇属 Mycena	11	5.26
				光柄菇属 Pluteus	5	2.39
				红菇属 Russula	17	8.13
				栓菌属 Trametes	5	2.39

关闭全屏

BibTeX
EndNote
RefWorks
TxT