Fault Diagnosis for Power Converters based on Random Forests and Feature Transformation

Lei Kou; Chuang Liu; Guo-wei Cai; Zhe Zhang; Xue-jiao Li; Quan-de Yuan

doi:10.1109/IPEMC-ECCEAsia48364.2020.9367970

Fault Diagnosis for Power Converters based on Random Forests and Feature Transformation

Lei Kou
, Chuang Liu
, Guo-wei Cai
, Zhe Zhang
, Xue-jiao Li
, Quan-de Yuan

Northeast Electric Power University
Harbin Institute of Technology
Changchun Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

313 Downloads (Orbit)

Abstract

Power converters have been key enablers of many energy conversion fields, and it is a trend to apply artificial intelligent (AI) technology to power converters to improve stability. A novel fault diagnosis method based on the combination of random forests (RFs) and feature transformation is proposed in this paper. Firstly, the three-phase AC fault currents of threephase PWM rectifier are analyzed as examples. Secondly, the feature transformation, a novel current trajectories slopes based method, is adopted to transform the fault currents data. With the help of feature transformation, the fault diagnosis classifier can obtain a good load adaptability. And then the RFs based method, a data-driven method, is employed to train the fault diagnosis
classifier with the fault current trajectories slopes data. Finally, the proposed method is carried out on an three-phase PWM rectifier system, which can detect and locate the open-circuit faults of IGBTs.

Original language	English
Title of host publication	Proceedings of 2020 IEEE 9^th International Power Electronics and Motion Control Conference
Publisher	IEEE
Publication date	2021
Pages	1821-1826
ISBN (Print)	978-1-7281-5302-5
DOIs	https://doi.org/10.1109/IPEMC-ECCEAsia48364.2020.9367970
Publication status	Published - 2021
Event	2020 IEEE 9^th International Power Electronics and Motion Control Conference - International Youth Cultural Centre, Nanjing, China Duration: 29 Nov 2020 → 2 Dec 2020 https://epe-ecce-conferences.com/epe2020/ipemc-2020-ecce-asia/

Conference

Conference	2020 IEEE 9^th International Power Electronics and Motion Control Conference
Location	International Youth Cultural Centre
Country/Territory	China
City	Nanjing
Period	29/11/2020 → 02/12/2020
Internet address	https://epe-ecce-conferences.com/epe2020/ipemc-2020-ecce-asia/

Keywords

Fault diagnosis
Power converters
Data-driven
Random forests
Feature transformation
Current trajectories slopes

Access to Document

10.1109/IPEMC-ECCEAsia48364.2020.9367970

FulltextAccepted author manuscript, 2.06 MB

OpenUrl availability

Check availability in DTU Findit

Cite this

@inproceedings{9a65457c49be4d81813ceb5cd10bf94b,

title = "Fault Diagnosis for Power Converters based on Random Forests and Feature Transformation",

abstract = "Power converters have been key enablers of many energy conversion fields, and it is a trend to apply artificial intelligent (AI) technology to power converters to improve stability. A novel fault diagnosis method based on the combination of random forests (RFs) and feature transformation is proposed in this paper. Firstly, the three-phase AC fault currents of threephase PWM rectifier are analyzed as examples. Secondly, the feature transformation, a novel current trajectories slopes based method, is adopted to transform the fault currents data. With the help of feature transformation, the fault diagnosis classifier can obtain a good load adaptability. And then the RFs based method, a data-driven method, is employed to train the fault diagnosis classifier with the fault current trajectories slopes data. Finally, the proposed method is carried out on an three-phase PWM rectifier system, which can detect and locate the open-circuit faults of IGBTs.",

keywords = "Fault diagnosis, Power converters, Data-driven, Random forests, Feature transformation, Current trajectories slopes",

author = "Lei Kou and Chuang Liu and Guo-wei Cai and Zhe Zhang and Xue-jiao Li and Quan-de Yuan",

year = "2021",

doi = "10.1109/IPEMC-ECCEAsia48364.2020.9367970",

language = "English",

isbn = "978-1-7281-5302-5",

pages = "1821--1826",

booktitle = "Proceedings of 2020 IEEE 9th International Power Electronics and Motion Control Conference",

publisher = "IEEE",

address = "United States",

note = "2020 IEEE 9<sup>th</sup> International Power Electronics and Motion Control Conference, IPEMC2020 ; Conference date: 29-11-2020 Through 02-12-2020",

url = "https://epe-ecce-conferences.com/epe2020/ipemc-2020-ecce-asia/",

}

Kou, L, Liu, C, Cai, G, Zhang, Z, Li, X & Yuan, Q 2021, Fault Diagnosis for Power Converters based on Random Forests and Feature Transformation. in Proceedings of 2020 IEEE 9^th International Power Electronics and Motion Control Conference. IEEE, pp. 1821-1826, 2020 IEEE 9^th International Power Electronics and Motion Control Conference, Nanjing, China, 29/11/2020. https://doi.org/10.1109/IPEMC-ECCEAsia48364.2020.9367970

TY - GEN

T1 - Fault Diagnosis for Power Converters based on Random Forests and Feature Transformation

AU - Kou, Lei

AU - Liu, Chuang

AU - Cai, Guo-wei

AU - Zhang, Zhe

AU - Li, Xue-jiao

AU - Yuan, Quan-de

PY - 2021

Y1 - 2021

N2 - Power converters have been key enablers of many energy conversion fields, and it is a trend to apply artificial intelligent (AI) technology to power converters to improve stability. A novel fault diagnosis method based on the combination of random forests (RFs) and feature transformation is proposed in this paper. Firstly, the three-phase AC fault currents of threephase PWM rectifier are analyzed as examples. Secondly, the feature transformation, a novel current trajectories slopes based method, is adopted to transform the fault currents data. With the help of feature transformation, the fault diagnosis classifier can obtain a good load adaptability. And then the RFs based method, a data-driven method, is employed to train the fault diagnosis classifier with the fault current trajectories slopes data. Finally, the proposed method is carried out on an three-phase PWM rectifier system, which can detect and locate the open-circuit faults of IGBTs.

AB - Power converters have been key enablers of many energy conversion fields, and it is a trend to apply artificial intelligent (AI) technology to power converters to improve stability. A novel fault diagnosis method based on the combination of random forests (RFs) and feature transformation is proposed in this paper. Firstly, the three-phase AC fault currents of threephase PWM rectifier are analyzed as examples. Secondly, the feature transformation, a novel current trajectories slopes based method, is adopted to transform the fault currents data. With the help of feature transformation, the fault diagnosis classifier can obtain a good load adaptability. And then the RFs based method, a data-driven method, is employed to train the fault diagnosis classifier with the fault current trajectories slopes data. Finally, the proposed method is carried out on an three-phase PWM rectifier system, which can detect and locate the open-circuit faults of IGBTs.

KW - Fault diagnosis

KW - Power converters

KW - Data-driven

KW - Random forests

KW - Feature transformation

KW - Current trajectories slopes

U2 - 10.1109/IPEMC-ECCEAsia48364.2020.9367970

DO - 10.1109/IPEMC-ECCEAsia48364.2020.9367970

M3 - Article in proceedings

SN - 978-1-7281-5302-5

SP - 1821

EP - 1826

BT - Proceedings of 2020 IEEE 9th International Power Electronics and Motion Control Conference

PB - IEEE

T2 - 2020 IEEE 9<sup>th</sup> International Power Electronics and Motion Control Conference

Y2 - 29 November 2020 through 2 December 2020

ER -

Fault Diagnosis for Power Converters based on Random Forests and Feature Transformation

Abstract

Conference

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this