Fault Diagnosis of Open Circuit Multiple IGBT's using PPCA-SVM in Single Phase Five-Level Voltage Controlled H-Bridge MLI

カテゴリ: 論文誌(論文単位)

グループ名: 【D】産業応用部門（英文）

発行日: 2020/01/01

タイトル(英語): Fault Diagnosis of Open Circuit Multiple IGBT's using PPCA-SVM in Single Phase Five-Level Voltage Controlled H-Bridge MLI

著者名: Nagendra Vara Prasad Kuraku (School of Electrical Engineering and Automation, Hefei University of Technology), Yigang He (School of Electrical Engineering and Automation, Hefei University of Technology), Murad Ali (School of Electrical Engineering and Aut

著者名(英語): Nagendra Vara Prasad Kuraku (School of Electrical Engineering and Automation, Hefei University of Technology), Yigang He (School of Electrical Engineering and Automation, Hefei University of Technology), Murad Ali (School of Electrical Engineering and Automation, Hefei University of Technology)

キーワード: fault diagnosis，voltage controlled CHMI，PPCA，SVM

要約(英語): Currently, multilevel inverters (MIs) are very popular in many industrial and renewable energy applications. Fast and accurate fault diagnosis is very important for improving their reliability. The present study proposes a fault diagnosis method based on the probabilistic principle component analysis (PPCA) and support vector machine (SVM) to control switches in single phase five-level voltage controlled cascaded H-Bridge MI (CHMI). The output voltage signals under different fault conditions of the CHMI are taken as fault features by using the phase-shift pulse width modulation technique. PPCA is used to optimize the data and reduce the dimension of the fault features. Finally, SVM classifier is used to diagnose the different fault modes. An experimental setup of CHMI is designed to validate the proposed fault diagnosis method. The simulation and experimental results show that by using PPCA-SVM, the accuracy of the fault location can be improved, and the time require for the fault diagnosis in CHMI can be reduced.

本誌: IEEJ Journal of Industry Applications Vol.9 No.1 （2020）

本誌掲載ページ: 61-72 p

原稿種別: 論文／英語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejjia/9/1/9_61/_article/-char/ja/