Deep and Machine Learning-based Methods for Defect Classification in Jet Engines

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Autoren

Externe Organisationen

  • Idaho State University
  • MTU Maintenance GmbH
  • Exzellenzcluster SE²A Sustainable and Energy-Efficient Aviation
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Titel des Sammelwerks2023 Intermountain Engineering, Technology and Computing, IETC 2023
Herausgeber (Verlag)Institute of Electrical and Electronics Engineers Inc.
Seiten43-48
Seitenumfang6
ISBN (elektronisch)9798350335903
ISBN (Print)979-8-3503-3591-0
PublikationsstatusVeröffentlicht - 2023
Veranstaltung2023 Annual Intermountain Engineering, Technology and Computing, IETC 2023 - Provo, USA / Vereinigte Staaten
Dauer: 12 Mai 202313 Mai 2023

Abstract

In this paper, the utility and accuracy of Machine Learning (ML) and Deep Learning (DL) methods are investigated for detecting defects in civil aircraft engines. Rather than to disassemble jet engines, the approach investigated in this study utilizes images of the exhaust of jet engines and infers defects in the turbine and burner section. While the proposed DL methods make use of one or two cameras, the ML methods depend on data obtained by extracting the density fields of the Hot Gas Path (HGP). The HPG data are computed from images acquired by an array of cameras. The corresponding ML features are crafted from these density fields. The proposed algorithms employ optimized hyperparameters and separate training as well as validation data sets. The study illustrates the potential of DL methods and the resulting simplification in the necessary instrumentation to accomplish near perfect defect classification outcomes.

ASJC Scopus Sachgebiete

Zitieren

Deep and Machine Learning-based Methods for Defect Classification in Jet Engines. / Schoen, Marco P.; Oettinger, Marcel; Mimic, Dajan.
2023 Intermountain Engineering, Technology and Computing, IETC 2023. Institute of Electrical and Electronics Engineers Inc., 2023. S. 43-48.

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Schoen, MP, Oettinger, M & Mimic, D 2023, Deep and Machine Learning-based Methods for Defect Classification in Jet Engines. in 2023 Intermountain Engineering, Technology and Computing, IETC 2023. Institute of Electrical and Electronics Engineers Inc., S. 43-48, 2023 Annual Intermountain Engineering, Technology and Computing, IETC 2023, Provo, USA / Vereinigte Staaten, 12 Mai 2023. https://doi.org/10.1109/IETC57902.2023.10152188
Schoen, M. P., Oettinger, M., & Mimic, D. (2023). Deep and Machine Learning-based Methods for Defect Classification in Jet Engines. In 2023 Intermountain Engineering, Technology and Computing, IETC 2023 (S. 43-48). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IETC57902.2023.10152188
Schoen MP, Oettinger M, Mimic D. Deep and Machine Learning-based Methods for Defect Classification in Jet Engines. in 2023 Intermountain Engineering, Technology and Computing, IETC 2023. Institute of Electrical and Electronics Engineers Inc. 2023. S. 43-48 doi: 10.1109/IETC57902.2023.10152188
Schoen, Marco P. ; Oettinger, Marcel ; Mimic, Dajan. / Deep and Machine Learning-based Methods for Defect Classification in Jet Engines. 2023 Intermountain Engineering, Technology and Computing, IETC 2023. Institute of Electrical and Electronics Engineers Inc., 2023. S. 43-48
Download
@inproceedings{12a1ebbf66f6447eba1bfc2bf20195e3,
title = "Deep and Machine Learning-based Methods for Defect Classification in Jet Engines",
abstract = "In this paper, the utility and accuracy of Machine Learning (ML) and Deep Learning (DL) methods are investigated for detecting defects in civil aircraft engines. Rather than to disassemble jet engines, the approach investigated in this study utilizes images of the exhaust of jet engines and infers defects in the turbine and burner section. While the proposed DL methods make use of one or two cameras, the ML methods depend on data obtained by extracting the density fields of the Hot Gas Path (HGP). The HPG data are computed from images acquired by an array of cameras. The corresponding ML features are crafted from these density fields. The proposed algorithms employ optimized hyperparameters and separate training as well as validation data sets. The study illustrates the potential of DL methods and the resulting simplification in the necessary instrumentation to accomplish near perfect defect classification outcomes.",
keywords = "Deep learning, defect classification, hot gas path, jet engines, machine learning",
author = "Schoen, {Marco P.} and Marcel Oettinger and Dajan Mimic",
note = "Funding Information: This research is based on data funded by the German Science Foundation (DFG) within the framework of the collaborative research center CRC 871 Regeneration of Complex Capital Goods funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) SFB 871/3 119193472. We would also like to acknowledge the funding by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany s Excellence Strategy EXC 2163/1 Sustainable and Energy Efficient Aviation Project ID 390881007. ; 2023 Annual Intermountain Engineering, Technology and Computing, IETC 2023 ; Conference date: 12-05-2023 Through 13-05-2023",
year = "2023",
doi = "10.1109/IETC57902.2023.10152188",
language = "English",
isbn = "979-8-3503-3591-0",
pages = "43--48",
booktitle = "2023 Intermountain Engineering, Technology and Computing, IETC 2023",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

Download

TY - GEN

T1 - Deep and Machine Learning-based Methods for Defect Classification in Jet Engines

AU - Schoen, Marco P.

AU - Oettinger, Marcel

AU - Mimic, Dajan

N1 - Funding Information: This research is based on data funded by the German Science Foundation (DFG) within the framework of the collaborative research center CRC 871 Regeneration of Complex Capital Goods funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) SFB 871/3 119193472. We would also like to acknowledge the funding by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany s Excellence Strategy EXC 2163/1 Sustainable and Energy Efficient Aviation Project ID 390881007.

PY - 2023

Y1 - 2023

N2 - In this paper, the utility and accuracy of Machine Learning (ML) and Deep Learning (DL) methods are investigated for detecting defects in civil aircraft engines. Rather than to disassemble jet engines, the approach investigated in this study utilizes images of the exhaust of jet engines and infers defects in the turbine and burner section. While the proposed DL methods make use of one or two cameras, the ML methods depend on data obtained by extracting the density fields of the Hot Gas Path (HGP). The HPG data are computed from images acquired by an array of cameras. The corresponding ML features are crafted from these density fields. The proposed algorithms employ optimized hyperparameters and separate training as well as validation data sets. The study illustrates the potential of DL methods and the resulting simplification in the necessary instrumentation to accomplish near perfect defect classification outcomes.

AB - In this paper, the utility and accuracy of Machine Learning (ML) and Deep Learning (DL) methods are investigated for detecting defects in civil aircraft engines. Rather than to disassemble jet engines, the approach investigated in this study utilizes images of the exhaust of jet engines and infers defects in the turbine and burner section. While the proposed DL methods make use of one or two cameras, the ML methods depend on data obtained by extracting the density fields of the Hot Gas Path (HGP). The HPG data are computed from images acquired by an array of cameras. The corresponding ML features are crafted from these density fields. The proposed algorithms employ optimized hyperparameters and separate training as well as validation data sets. The study illustrates the potential of DL methods and the resulting simplification in the necessary instrumentation to accomplish near perfect defect classification outcomes.

KW - Deep learning

KW - defect classification

KW - hot gas path

KW - jet engines

KW - machine learning

UR - http://www.scopus.com/inward/record.url?scp=85164539539&partnerID=8YFLogxK

U2 - 10.1109/IETC57902.2023.10152188

DO - 10.1109/IETC57902.2023.10152188

M3 - Conference contribution

AN - SCOPUS:85164539539

SN - 979-8-3503-3591-0

SP - 43

EP - 48

BT - 2023 Intermountain Engineering, Technology and Computing, IETC 2023

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2023 Annual Intermountain Engineering, Technology and Computing, IETC 2023

Y2 - 12 May 2023 through 13 May 2023

ER -

Von denselben Autoren