Non-stationary condition monitoring through event alignment

Niels Henrik Pontoppidan; Jan Larsen

doi:10.1109/MLSP.2004.1423012

Non-stationary condition monitoring through event alignment

Niels Henrik Pontoppidan, Jan Larsen

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

348 Downloads (Pure)

Abstract

We present an event alignment framework which enables change detection in non-stationary signals. change detection. Classical condition monitoring frameworks have been restrained to laboratory settings with stationary operating conditions, which are not resembling real world operation. In this paper we apply the technique for non-stationary condition monitoring of large diesel engines based on acoustical emission sensor signals. The performance of the event alignment is analyzed in an unsupervised probabilistic detection framework based on outlier detection with either Principal Component Analysis or Gaussian Processes modeling. We are especially interested in the true performance of the condition monitoring performance with mixed aligned and unaligned data, e.g. detection of fault condition of unaligned examples versus false alarms of aligned normal condition data. Further, we expect that the non-stationary model can be used for wear trending due to longer and continuous monitoring across operating condition changes.

Original language	English
Title of host publication	IEEE Workshop on Machine Learning for Signal Processing
Publisher	IEEE Press
Publication date	2004
Pages	499-508
ISBN (Print)	0-7803-8608-4
DOIs	https://doi.org/10.1109/MLSP.2004.1423012
Publication status	Published - 2004
Event	2004 14th IEEE Workshop on Machine Learning for Signal Processing - São Luis, Brazil Duration: 29 Sept 2004 → 1 Oct 2004 Conference number: 14 https://ieeexplore.ieee.org/xpl/conhome/9735/proceeding

Conference

Conference	2004 14th IEEE Workshop on Machine Learning for Signal Processing
Number	14
Country/Territory	Brazil
City	São Luis
Period	29/09/2004 → 01/10/2004
Internet address	https://ieeexplore.ieee.org/xpl/conhome/9735/proceeding

Bibliographical note

Copyright: 2004 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE

Keywords

Condition Monitoring
Non-stationarity
Diesel Engine

Access to Document

10.1109/MLSP.2004.1423012

Imm3131Final published version, 615 KB

OpenUrl availability

Full text

Cite this

@inproceedings{14c4ed5bba034d5499ab06acdc36d6e9,

title = "Non-stationary condition monitoring through event alignment",

abstract = "We present an event alignment framework which enables change detection in non-stationary signals. change detection. Classical condition monitoring frameworks have been restrained to laboratory settings with stationary operating conditions, which are not resembling real world operation. In this paper we apply the technique for non-stationary condition monitoring of large diesel engines based on acoustical emission sensor signals. The performance of the event alignment is analyzed in an unsupervised probabilistic detection framework based on outlier detection with either Principal Component Analysis or Gaussian Processes modeling. We are especially interested in the true performance of the condition monitoring performance with mixed aligned and unaligned data, e.g. detection of fault condition of unaligned examples versus false alarms of aligned normal condition data. Further, we expect that the non-stationary model can be used for wear trending due to longer and continuous monitoring across operating condition changes.",

keywords = "Condition Monitoring, Non-stationarity, Diesel Engine",

author = "Pontoppidan, {Niels Henrik} and Jan Larsen",

note = "Copyright: 2004 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE; 2004 14th IEEE Workshop on Machine Learning for Signal Processing, MLSP 2004 ; Conference date: 29-09-2004 Through 01-10-2004",

year = "2004",

doi = "10.1109/MLSP.2004.1423012",

language = "English",

isbn = "0-7803-8608-4",

pages = "499--508",

booktitle = "IEEE Workshop on Machine Learning for Signal Processing",

publisher = "IEEE Press",

url = "https://ieeexplore.ieee.org/xpl/conhome/9735/proceeding",

}

TY - GEN

T1 - Non-stationary condition monitoring through event alignment

AU - Pontoppidan, Niels Henrik

AU - Larsen, Jan

N1 - Conference code: 14

PY - 2004

Y1 - 2004

N2 - We present an event alignment framework which enables change detection in non-stationary signals. change detection. Classical condition monitoring frameworks have been restrained to laboratory settings with stationary operating conditions, which are not resembling real world operation. In this paper we apply the technique for non-stationary condition monitoring of large diesel engines based on acoustical emission sensor signals. The performance of the event alignment is analyzed in an unsupervised probabilistic detection framework based on outlier detection with either Principal Component Analysis or Gaussian Processes modeling. We are especially interested in the true performance of the condition monitoring performance with mixed aligned and unaligned data, e.g. detection of fault condition of unaligned examples versus false alarms of aligned normal condition data. Further, we expect that the non-stationary model can be used for wear trending due to longer and continuous monitoring across operating condition changes.

AB - We present an event alignment framework which enables change detection in non-stationary signals. change detection. Classical condition monitoring frameworks have been restrained to laboratory settings with stationary operating conditions, which are not resembling real world operation. In this paper we apply the technique for non-stationary condition monitoring of large diesel engines based on acoustical emission sensor signals. The performance of the event alignment is analyzed in an unsupervised probabilistic detection framework based on outlier detection with either Principal Component Analysis or Gaussian Processes modeling. We are especially interested in the true performance of the condition monitoring performance with mixed aligned and unaligned data, e.g. detection of fault condition of unaligned examples versus false alarms of aligned normal condition data. Further, we expect that the non-stationary model can be used for wear trending due to longer and continuous monitoring across operating condition changes.

KW - Condition Monitoring

KW - Non-stationarity

KW - Diesel Engine

U2 - 10.1109/MLSP.2004.1423012

DO - 10.1109/MLSP.2004.1423012

M3 - Article in proceedings

SN - 0-7803-8608-4

SP - 499

EP - 508

BT - IEEE Workshop on Machine Learning for Signal Processing

PB - IEEE Press

T2 - 2004 14th IEEE Workshop on Machine Learning for Signal Processing

Y2 - 29 September 2004 through 1 October 2004

ER -

Non-stationary condition monitoring through event alignment

Abstract

Conference

Bibliographical note

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this