Operation Characteristics Prediction Method of Press-pack IGBT Based on Deep Learning

XIAO Kai; WANG Zhen; ZOU Yansheng; YAN Xilin; XU Linhao; HUANG Chenhao

doi:10.13296/j.1001-1609.hva.2026.10.026

您当前的位置：

首页 >

文章列表页 >

Operation Characteristics Prediction Method of Press-pack IGBT Based on Deep Learning

Technical Discussion | 更新时间：2026-10-14

- Operation Characteristics Prediction Method of Press-pack IGBT Based on Deep Learning
- High Voltage Apparatus Vol. 62, Issue 10, Pages: 227-236(2026)
- 作者机构：
  
  中国南方电网有限责任公司超高压输电公司电力科研院，广州 510663
  湖南大学电气与信息工程学院，长沙 410082
- 作者简介：
- 基金信息：
- DOI：10.13296/j.1001-1609.hva.2026.10.026
  CLC：
- Received：22 May 2026，
  
  Revised：2026-07-10，
  
  Published：16 October 2026
- 稿件说明：
移动端阅览
XIAO Kai, WANG Zhen, ZOU Yansheng, et al. Operation Characteristics Prediction Method of Press-pack IGBT Based on Deep Learning[J]. High Voltage Apparatus, 2026, 62(10): 227-236.
DOI：

XIAO Kai, WANG Zhen, ZOU Yansheng, et al. Operation Characteristics Prediction Method of Press-pack IGBT Based on Deep Learning[J]. High Voltage Apparatus, 2026, 62(10): 227-236. DOI： 10.13296/j.1001-1609.hva.2026.10.026.

摘要

压接式IGBT是柔直输电系统的关键环境，其寿命评估是提高系统可靠性的重要环节，其中掌握IGBT模块多场特征是关键。尽管现有多场仿真技术能够实现压接式IGBT模块的完备分析，但是存在计算效率低的问题。因此，文中提出了融合深度学习和多场模型的压接式IGBT模块多场运行特征分析方法，以实现压接IGBT模块运行信息的快速获取，支撑其数字孪生模型的构建。首先，构建了压接式IGBT模块电—热—力多场耦合模型，分析了不同工况下IGBT模型重要寿命评估特征变化规律；其次，提出了BPBR神经网络与网格搜索法相结合的深度学习方法，以实现压接式IGBT模块多场耦合信息的快速预测；最后，测试了所提出方法的高效性，结果表明：所提出模型能够有效分析压接式IGBT模块多场信息，与有限元计算结果吻合度好，偏差率小于1%，且计算时间不到仿真的万分之一。研究成果能够支撑压接式IGBT模块多场信息的快速获取，进而支撑压接式IGBT模块数字孪生体的构建和状态实时监测。

Abstract

Press-pack IGBT is the key environment of VSC-HVDC transmission system

and its life assessment is a crucial part of improving system reliability

in which mastering the multi-field characteristics of IGBT module is of critical importance. Although the existing multi-field simulation technology can achieve thorough analysis of IGBT module

its computational efficiency remains limited. Therefore

in this paper an analysis method of multi-field operation characteristics of IGBT module with deep learning and multi-field model is proposed in order to realize the rapid acquisition of operation information of IGBT module and support the construction of its digital twin model. First

the electric-thermal-mechanical multi-field coupling model of press-pack IGBT module is constructed

and the variation of important life assessment characteristics of IGBT model under different working conditions is analyzed. Then

a deep learning method combining BPBR neural network and grid search method is proposed to realize the rapid prediction of multi-field coupling information of IGBT module. Finally

the high efficiency of the proposed method is tested. The result shows that the proposed model can effectively analyze the multi-field information of the IGBT module

which is in good agreement with the finite element calculation results

with deviation rate less than 1% and calculation time less than one ten-thousandth of that of the simulation. The research results can support the rapid acquisition of multi-field information of IGBT module

and then support the construction and real-time monitoring of digital twins of IGBT module.

关键词

Keywords

references

郑全旭,孙小平,娄彦涛,等.±500 kV海上柔直换流阀轻型化设计及仿真分析[J].高压电器,2023,59(6):188-201.

Zheng Quanxu,Sun Xiaoping,Lou Yantao,et al.Lightweight design and simulation analysis of ±500 kV offshore VSC converter valve[J].High Voltage Apparatus,2023,59(6):188-201.

贾科,刘可歆,毕天姝,等.基于波形匹配的柔性直流输电线路单端量保护研究[J].电力系统保护与控制,2025,53(3):13-26.

Jia Ke,Liu Kexin,Bi Tianshu,et al.Research on single-ended protection of a flexible HVDC transmission line based on waveform matching[J].Power System Protection and Control ,2025,53(3):13-26.

孙凯,卢世蕾,易哲嫄,等.面向电力电子变压器应用的大容量高频变压器技术综述[J].中国电机工程学报,2021,41(24):8531-8545.

Sun Kai,Lu Shilei,Yi Zheyuan,et al.A review of high-power highfrequency transformer technology for power electronic transformer applications[J].Proceedings of the CSEE,2021,41(24):8531-8545.

Han Lubin,Liang Lin,Wang Yijian,et al. Performance limits of high voltage press-pack SiC IGBT and SiC MOSFET devices[J].Power Electronic Devices and Components,2022(3):100019.

Oh H,Han B,Mccluskey P,et al.Physics-of-failure,condition monitoring,and prognostics of insulated gate bipolar transistor modules:a review[J].IEEE Transactions on Power Electronics, 2015,30(5):2413-2426.

Huang Zhiliang,Yang Tiangui,Li Chuanbo,et al.Optimization of the pressure distribution in press-pack insulated gate bipolar transistors[J].Structural and Multidisciplinary Optimization,2021, 63(2):855-865.

杨兵,王胜利,王文杰,等.混合式直流断路器IGBT关断能力提升技术研究[J].高压电器,2020,56(5):24-32.

Yang Bing,Wang Shengli,Wang Wenjie,et al.Research on IGBT turning off capacity enhancement technology of hybrid DC circuit breaker[J].High Voltage Apparatus,2020,56(5):24-32.

Liu Renkuan,Li Hui,Yao Ran,et al.Research on long-term reliability of silver sintered press-pack IGBT modules[J].Power Electronic Devices and Components,2022 (3):100012.

Lutz J,Schlangenotto H,Scheyermann U,et al.Semiconductor power devices:physics,characteristics,reliability[M].Berlin Heidelberg:Springer-Verlag,2011.

任斌,黄永章.压接型IGBT器件单芯片子模组可靠性寿命仿真研究[J].华北电力大学学报(自然科学版),2020,47(3):49-57.

Ren Bin,Huang Yongzhang.Simulation on reliability lifetime of single IGBT chip module of press-pack IGBTs[J].Journal of North China Electric Power University(Natural Science Edition),2020,47(3):49-57.

张经纬.基于有限元法的压接型IGBT器件单芯片子模组疲劳寿命预测[D].北京:华北电力大学,2018.

Zhang Jingwei.Fatigue life prediction of singl-echip submodule of press-pack IGBTs based on finite element method[D].Beijing:North China Electric Power University,2018.

郭佳奇,张一鸣,邓二平,等.考虑接触界面全接触的压接型IGBT器件功率循环寿命预测[J].华北电力大学学报(自然科学版),2023,50(2):63-72.

Guo Jiaqi,Zhang Yiming,Deng Erping,et al.Prediction of power cycle life of press-pack IGBTs considering full contact of contact interface[J].Journal of North China Electric Power University (Natural Science Edition),2023,50(2):63-72.

任斌.压接型IGBT器件可靠性寿命模型与相关仿真研究[D].北京:华北电力大学,2020.

Ren Bin.Simulation on reliability lifetime of single IGBT chip module of press-pack IGBTs[D].Beijing:North China Electric Power University,2020.

韩文超,路茂增,马新喜,等.高压大容量混合型MMC半桥子模块下部IGBT损耗优化方法[J].电力工程技术,2024,43(2):189-198.

Han Wenchao,Lu Maozeng,Ma Xinxi,et al.Loss optimization method for bottom IGBT in half bridge sub-module of high voltage and large capacity hybridMMC[J].Electric Power Engineering Technology,2024,43(2):189-198.

邓吉利.柔性直流换流阀压接式IGBT器件可靠性建模与评估[D].重庆:重庆大学,2018.

Deng Jili.Reliability modeling and evaluation of press-pack IGBT devices forVSC-HVDCconverter valves[D].Chongqing:Chongqing University,2018.

邓真宇,陈民铀,赖伟,等.多芯片并联压接式IGBT中温度不均对电流分布的影响[J].中国电力,2020,53(12):10-17.

Deng Zhenyu,Chen Minyou,Lai Wei,et al.Influence of uneven temperature on current distribution in paralleled multi-chips press packIGBT[J].Electric Power,2020,53(12):10-17.

李洁,党晓圆,邓真宇,等.多芯片并联压接式IGBT器件多物理场耦合建模分析[J].电力电子技术,2021,55(12):9-12.

Li Jie,Dang Xiaoyuan,Deng Zhenyu,et al.Multi-physical field coupling modeling analysis of multi-chip parallel press pack IGBT devices[J].Power Electronics,2021,55(12):9-12.

张筱迪,毛明晖,卢昶衡,等.基于有限元分析和机器学习的跌落所致封装结构力学行为预测[J].电子与封装,2021,21 (2):73-80.

Zhang Xiaodi,Mao Minghui,Lu Changheng,et al.Prediction of mechanical behavior of package structure subjected to drop impact based on finite element analysis and machine learning[J].Electronics and Packaging,2021,21(2):73-80.

金亮,王飞,杨庆新,等.永磁同步电机性能分析的典型深度学习模型与训练方法[J].电工技术学报,2018,33(s1):41-48.

Jin Liang,Wang Fei,Yang Qingxin,et al.Typical deep learning model and training method for performance analysis of permanent magnet synchronous motor[J].Transactions of China Electrotechnical Society,2018,33(s1):41-48.

王要强,陈晨,张世达,等.基于新型趋近律的表贴式永磁同步电机抗扰滑模控制策略[J].电力系统保护与控制,2026,54(2):151-161.

Wang Yaoqiang,Chen Chen,Zhang Shida,et al.Anti-disturbance sliding mode control strategy for surface - mounted permanent magnet synchronous motors based on a novel reaching law[J].Power System Protection and Control,2026,54(2):151-161.

魏尧,付俊荣,王高林,等.基于陷波抗停滞的永磁同步电机无模型预测电流滑模控制[J].电工技术学报,2026,41(2):475-486.

Wei Yao,Fu Junrong,Wang Gaolin,et al.Notch anti-stagnation based model - free predictive current sliding mode control of permanent magnet synchronous motor drives[J].Transactions of China Electrotechnical Society,2026,41(2):475-486.

李耀华,种国臣,王钦政,等.永磁同步电机模型预测电流控制简化研究[J].电力电子技术,2026,60(2):137-143.

Li Yaohua,Chong Guochen,Wang Qinzheng,et al.Research on simplified model predictive current control for permanent magnet synchronous motor[J].Power Electronics,2026,60(2):137-143.

宋诗雨,吴永斌,尹海韬,等.基于混合式BPNN/PSO-MRAS的永磁同步电机参数在线辨识[J].智慧电力,2026,54(4):111-118.

Song Shiyu,Wu Yongbin,Yin Haitao,et al.Online parameter identification of permanent magnet synchronous motor based on hybridBPNN/PSO-MRAS[J].Smart Power,2026,54(4):111-118.

王鸣宇,汤毅,许彦祺,等.表贴式永磁同步电机空载电磁性能解析分析[J].大电机技术,2026(1):40-48.

Wang Mingyu,Tang Yi,Xu Yanqi,et al. Analytical analysis of noload electromagnetic performance of surface mounted permanent magnet synchronous motor[J].Large Electric Machine and Hydraulic Turbine,2026(1):40-48.

刘向贞.永磁同步电机性能分析的深度学习算法与应用[D].天津:天津工业大学,2020.

Liu Xiangzhen.Deep learning algorithm for performance analysis of permanent magnet synchronous motor and its application[D].Tianjin:Tiangong University,2020.

阎世梁,王银玲,路丹丹,等.基于深度神经网络的永磁直线电机仿真与优化[J].系统仿真学报,2024,36(3):713-725.

Yan Shiliang,Wang Yinling,Lu Dandan,et al.Simulation and optimization of permanent magnet linear machine based on deep neural network[J].Journal of System Simulation ,2024 ,36(3) :713-725.

Jing Honglei. Deep neural network based chiller energy saving control method[J].Computer Informatization and Mechanical System,2023,6(5):73-78.

El-dalahmeh M ,Al-greer M ,El-dalahmeh M ,et al. Capacity estimation of lithium-ion batteries based on adaptive empirical wavelet transform and long short-term memory neural network[J].Journal of Energy Storage,2023 (70):108046.

Zhao Meiling,Zhang Yunwei,Yuan Zhanbin,et al.A point-to-point convolutional neural network for reconstructing electromagnetic parameters of multiple cavities scattering with inhomogeneous anisotropic media[J].Engineering Analysis With Boundary Elements,2023(155):281-296.

耿学锋,何赟泽,李孟川,等.IGBT多物理场建模技术与应用研究概述[J].中国电机工程学报,2022,42(1):271-290.

Geng Xuefeng,He Yunze,Li Mengchuan,et al.An overview of IGBT multiphysics modeling technology and application[J].Proceedings of the CSEE,2022,42(1):271-290.

范继锋,王永强,张天龙,等.基于时频分析矩阵与BP神经网络的输电线路保护方法[J].电力科学与技术学报,2026,41(1):46-52.

Fan Jifeng,Wang Yongqiang,Zhang Tianlong,et al.Transmission line protection method based on time-frequency analysismatrix and BP neural network[J].Journal of Electric Power Science and Technology,2026,41(1):46-52.

夏世远,苏建徽,杜燕,等.基于贝叶斯正则化BP神经网络的PEMFC电堆建模[J].合肥工业大学学报(自然科学版),2021,44 (7):895-899.

Xia Shiyuan,Su Jianhui,Du Yan,et al. Modeling of PEMFC stack based on Bayesian regularized BP neural network[J].Journal of Hefei University of Technology(Natural Science),2021,44(7):895-899.

Ghahramani Z.Probabilistic machine learning and artificial intelligence[J].Nature,2015,521(7553):452-459.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Distribution Characteristics of Electromagnetic Forces in Windings of Thyristor-controlled Phase Shifters Under Typical Fault Conditions

Research on Mechanical-electrical Collaborative Design of UHV Converter Valves Considering Seismic Performance

Partial Discharge Pattern Recognition of High Voltage Cables Based on SDAE-SVM

Study on Simulation Method for Temperature Distribution of High-voltage Cables Laid in Tunnels Based on Electric-thermal-fluid Fields Coupling

Relation-oriented Method for Jointly Extracing Text and Relations of Power Equipment Fault Defect Based on Relationships

Related Author

CAI Xiang

YAO Yongle

LI Qun

CHEN Shi

WANG Tonglei

PANG Lei

ZHAO Weijie

CUI Zhanglun

Related Institution

School of Electrical Engineering, Xi'an Jiaotong University

State Grid Jiangsu Electric Power Company Research Institute

Kunming Bureau of CSG EHV Transmission Company

Yunnan Provincial Engineering Research Center of UHVDC Smart O & M and Safe Operation

School of Electric Power Engineering in Kunming University of Science and Technology

⁰