Data-driven fault detection in aircraft engines with noisy sensor measurements

Soumik Sarkar, Xin Jin, Asok Ray

Research output: Contribution to journalArticle

45 Citations (Scopus)

Abstract

An inherent difficulty in sensor-data-driven fault detection is that the detection performance could be drastically reduced under sensor degradation (e.g., drift and noise). Complementary to traditional model-based techniques for fault detection, this paper proposes symbolic dynamic filtering by optimally partitioning the time series data of sensor observation. The objective here is to mask the effects of sensor noise level variation and magnify the system fault signatures. In this regard, the concepts of feature extraction and pattern classification are used for fault detection in aircraft gas turbine engines. The proposed methodology of data-driven fault detection is tested and validated on the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) test-bed developed by NASA for noisy (i.e., increased variance) sensor signals.

Original languageEnglish (US)
Article number081602
JournalJournal of Engineering for Gas Turbines and Power
Volume133
Issue number8
DOIs
StatePublished - Apr 21 2011

Fingerprint

Aircraft engines
Fault detection
Sensors
Propulsion
Pattern recognition
Gas turbines
NASA
Feature extraction
Masks
Time series
Turbines
Aircraft
Degradation

All Science Journal Classification (ASJC) codes

  • Nuclear Energy and Engineering
  • Fuel Technology
  • Aerospace Engineering
  • Energy Engineering and Power Technology
  • Mechanical Engineering

Cite this

@article{217433a87d544ad788f45368ff3a7523,
title = "Data-driven fault detection in aircraft engines with noisy sensor measurements",
abstract = "An inherent difficulty in sensor-data-driven fault detection is that the detection performance could be drastically reduced under sensor degradation (e.g., drift and noise). Complementary to traditional model-based techniques for fault detection, this paper proposes symbolic dynamic filtering by optimally partitioning the time series data of sensor observation. The objective here is to mask the effects of sensor noise level variation and magnify the system fault signatures. In this regard, the concepts of feature extraction and pattern classification are used for fault detection in aircraft gas turbine engines. The proposed methodology of data-driven fault detection is tested and validated on the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) test-bed developed by NASA for noisy (i.e., increased variance) sensor signals.",
author = "Soumik Sarkar and Xin Jin and Asok Ray",
year = "2011",
month = "4",
day = "21",
doi = "10.1115/1.4002877",
language = "English (US)",
volume = "133",
journal = "Journal of Engineering for Gas Turbines and Power",
issn = "0742-4795",
publisher = "American Society of Mechanical Engineers(ASME)",
number = "8",

}

Data-driven fault detection in aircraft engines with noisy sensor measurements. / Sarkar, Soumik; Jin, Xin; Ray, Asok.

In: Journal of Engineering for Gas Turbines and Power, Vol. 133, No. 8, 081602, 21.04.2011.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Data-driven fault detection in aircraft engines with noisy sensor measurements

AU - Sarkar, Soumik

AU - Jin, Xin

AU - Ray, Asok

PY - 2011/4/21

Y1 - 2011/4/21

N2 - An inherent difficulty in sensor-data-driven fault detection is that the detection performance could be drastically reduced under sensor degradation (e.g., drift and noise). Complementary to traditional model-based techniques for fault detection, this paper proposes symbolic dynamic filtering by optimally partitioning the time series data of sensor observation. The objective here is to mask the effects of sensor noise level variation and magnify the system fault signatures. In this regard, the concepts of feature extraction and pattern classification are used for fault detection in aircraft gas turbine engines. The proposed methodology of data-driven fault detection is tested and validated on the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) test-bed developed by NASA for noisy (i.e., increased variance) sensor signals.

AB - An inherent difficulty in sensor-data-driven fault detection is that the detection performance could be drastically reduced under sensor degradation (e.g., drift and noise). Complementary to traditional model-based techniques for fault detection, this paper proposes symbolic dynamic filtering by optimally partitioning the time series data of sensor observation. The objective here is to mask the effects of sensor noise level variation and magnify the system fault signatures. In this regard, the concepts of feature extraction and pattern classification are used for fault detection in aircraft gas turbine engines. The proposed methodology of data-driven fault detection is tested and validated on the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) test-bed developed by NASA for noisy (i.e., increased variance) sensor signals.

UR - http://www.scopus.com/inward/record.url?scp=79954507840&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79954507840&partnerID=8YFLogxK

U2 - 10.1115/1.4002877

DO - 10.1115/1.4002877

M3 - Article

AN - SCOPUS:79954507840

VL - 133

JO - Journal of Engineering for Gas Turbines and Power

JF - Journal of Engineering for Gas Turbines and Power

SN - 0742-4795

IS - 8

M1 - 081602

ER -