Research on Transformer Based Online Condition Monitoring Method for IPM

IF 2.7 3区工程技术 Q1 ENGINEERING, MULTIDISCIPLINARY Measurement Science and Technology Pub Date : 2024-07-15 DOI:10.1088/1361-6501/ad6341

Shengxue Tang, Jinze Zhao, Liqiang Tan, Jinjing Yan

{"title":"Research on Transformer Based Online Condition Monitoring Method for IPM","authors":"Shengxue Tang, Jinze Zhao, Liqiang Tan, Jinjing Yan","doi":"10.1088/1361-6501/ad6341","DOIUrl":null,"url":null,"abstract":"\n The Intelligent Power Module (IPM) has the integrated packaging, leading to the advantages of convenient use, safety and reliability. However, once it fails, it will cause the whole power supply system to be inoperative, and it is necessary to perform online Condition Monitoring (CM) of the IPM. In this paper, we extract the aging characteristics such as dynamic equivalent resistance, peak-to-peak value, switching frequency, and turn-off time only from the voltage and current of IPM drain-source port, and then propose a non-intrusive online CM method for the IPM based on the Transformer Neural Network (TNN). We analyse the internal aging mechanism of the IPM, the changing law of aging features, and construct multi-dimensional aging fusion features, and then the TNN model is used to monitor early parameters drift of multi-dimensional fusion feature vectors and realize the accurate online prediction of IPM health condition. The experimental analysis results show that the fault prediction accuracy reaches 96%, and can realize the health CM under the condition of noise interference, weak aging features and few external observable points.","PeriodicalId":18526,"journal":{"name":"Measurement Science and Technology","volume":null,"pages":null},"PeriodicalIF":2.7000,"publicationDate":"2024-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Measurement Science and Technology","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1088/1361-6501/ad6341","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, MULTIDISCIPLINARY","Score":null,"Total":0}

引用次数: 0

Abstract

The Intelligent Power Module (IPM) has the integrated packaging, leading to the advantages of convenient use, safety and reliability. However, once it fails, it will cause the whole power supply system to be inoperative, and it is necessary to perform online Condition Monitoring (CM) of the IPM. In this paper, we extract the aging characteristics such as dynamic equivalent resistance, peak-to-peak value, switching frequency, and turn-off time only from the voltage and current of IPM drain-source port, and then propose a non-intrusive online CM method for the IPM based on the Transformer Neural Network (TNN). We analyse the internal aging mechanism of the IPM, the changing law of aging features, and construct multi-dimensional aging fusion features, and then the TNN model is used to monitor early parameters drift of multi-dimensional fusion feature vectors and realize the accurate online prediction of IPM health condition. The experimental analysis results show that the fault prediction accuracy reaches 96%, and can realize the health CM under the condition of noise interference, weak aging features and few external observable points.

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

基于变压器的 IPM 在线状态监测方法研究

智能电源模块（IPM）采用一体化封装，具有使用方便、安全可靠等优点。但是，一旦出现故障，就会导致整个供电系统无法工作，因此有必要对 IPM 进行在线状态监测（CM）。本文仅从 IPM 漏源端口的电压和电流中提取动态等效电阻、峰峰值、开关频率和关断时间等老化特性，然后提出一种基于变压器神经网络（TNN）的 IPM 非侵入式在线 CM 方法。我们分析了 IPM 内部老化机理、老化特征变化规律，构建了多维老化融合特征，然后利用 TNN 模型监测多维融合特征向量的早期参数漂移，实现了对 IPM 健康状况的精确在线预测。实验分析结果表明，故障预测准确率达到 96%，能够在噪声干扰、老化特征弱、外部观测点少的情况下实现健康 CM。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

Measurement Science and Technology 工程技术-工程：综合

CiteScore

4.30

自引率

16.70%

发文量

656

审稿时长

4.9 months

期刊介绍： Measurement Science and Technology publishes articles on new measurement techniques and associated instrumentation. Papers that describe experiments must represent an advance in measurement science or measurement technique rather than the application of established experimental technique. Bearing in mind the multidisciplinary nature of the journal, authors must provide an introduction to their work that makes clear the novelty, significance, broader relevance of their work in a measurement context and relevance to the readership of Measurement Science and Technology. All submitted articles should contain consideration of the uncertainty, precision and/or accuracy of the measurements presented. Subject coverage includes the theory, practice and application of measurement in physics, chemistry, engineering and the environmental and life sciences from inception to commercial exploitation. Publications in the journal should emphasize the novelty of reported methods, characterize them and demonstrate their performance using examples or applications.