Helicopter vibration sensor selection using data visualisation

Waljinder S. Gill, Ian T. Nabney, D. Wells

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The main objective of the project is to enhance the already effective health-monitoring system (HUMS) for helicopters by analysing structural vibrations to recognise different flight conditions directly from sensor information. The goal of this paper is to develop a new method to select those sensors and frequency bands that are best for detecting changes in flight conditions. We projected frequency information to a 2-dimensional space in order to visualise flight-condition transitions using the Generative Topographic Mapping (GTM) and a variant which supports simultaneous feature selection. We created an objective measure of the separation between different flight conditions in the visualisation space by calculating the Kullback-Leibler (KL) divergence between Gaussian mixture models (GMMs) fitted to each class: the higher the KL-divergence, the better the interclass separation. To find the optimal combination of sensors, they were considered in pairs, triples and groups of four sensors. The sensor triples provided the best result in terms of KL-divergence. We also found that the use of a variational training algorithm for the GMMs gave more reliable results.
Original languageEnglish
Title of host publicationIEEE International Workshop on Machine Learning for Signal Processing, MLSP 2012
PublisherIEEE
Number of pages6
ISBN (Electronic)978-1-4673-1025-3
ISBN (Print)978-1-4673-1024-6
DOIs
Publication statusPublished - Jul 2012
EventIEEE International Workshop on Machine Learning for Signal Processing (MLSP 2012) - Santander, Spain
Duration: 23 Sep 201226 Sep 2012

Publication series

NameMachine learning for signal processing
PublisherIEEE
ISSN (Print)1551-2541

Conference

ConferenceIEEE International Workshop on Machine Learning for Signal Processing (MLSP 2012)
CountrySpain
CitySantander
Period23/09/1226/09/12

Fingerprint

Data visualization
Helicopters
Sensors
Frequency bands
Feature extraction
Visualization
Health
Monitoring

Bibliographical note

© 2012 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Cite this

Gill, W. S., Nabney, I. T., & Wells, D. (2012). Helicopter vibration sensor selection using data visualisation. In IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2012 (Machine learning for signal processing). IEEE. https://doi.org/10.1109/MLSP.2012.6349808
Gill, Waljinder S. ; Nabney, Ian T. ; Wells, D. / Helicopter vibration sensor selection using data visualisation. IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2012. IEEE, 2012. (Machine learning for signal processing).
@inproceedings{caa2749bc9a044bca4d592cfb8ad412f,
title = "Helicopter vibration sensor selection using data visualisation",
abstract = "The main objective of the project is to enhance the already effective health-monitoring system (HUMS) for helicopters by analysing structural vibrations to recognise different flight conditions directly from sensor information. The goal of this paper is to develop a new method to select those sensors and frequency bands that are best for detecting changes in flight conditions. We projected frequency information to a 2-dimensional space in order to visualise flight-condition transitions using the Generative Topographic Mapping (GTM) and a variant which supports simultaneous feature selection. We created an objective measure of the separation between different flight conditions in the visualisation space by calculating the Kullback-Leibler (KL) divergence between Gaussian mixture models (GMMs) fitted to each class: the higher the KL-divergence, the better the interclass separation. To find the optimal combination of sensors, they were considered in pairs, triples and groups of four sensors. The sensor triples provided the best result in terms of KL-divergence. We also found that the use of a variational training algorithm for the GMMs gave more reliable results.",
author = "Gill, {Waljinder S.} and Nabney, {Ian T.} and D. Wells",
note = "{\circledC} 2012 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.",
year = "2012",
month = "7",
doi = "10.1109/MLSP.2012.6349808",
language = "English",
isbn = "978-1-4673-1024-6",
series = "Machine learning for signal processing",
publisher = "IEEE",
booktitle = "IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2012",
address = "United States",

}

Gill, WS, Nabney, IT & Wells, D 2012, Helicopter vibration sensor selection using data visualisation. in IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2012. Machine learning for signal processing, IEEE, IEEE International Workshop on Machine Learning for Signal Processing (MLSP 2012), Santander, Spain, 23/09/12. https://doi.org/10.1109/MLSP.2012.6349808

Helicopter vibration sensor selection using data visualisation. / Gill, Waljinder S.; Nabney, Ian T.; Wells, D.

IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2012. IEEE, 2012. (Machine learning for signal processing).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Helicopter vibration sensor selection using data visualisation

AU - Gill, Waljinder S.

AU - Nabney, Ian T.

AU - Wells, D.

N1 - © 2012 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

PY - 2012/7

Y1 - 2012/7

N2 - The main objective of the project is to enhance the already effective health-monitoring system (HUMS) for helicopters by analysing structural vibrations to recognise different flight conditions directly from sensor information. The goal of this paper is to develop a new method to select those sensors and frequency bands that are best for detecting changes in flight conditions. We projected frequency information to a 2-dimensional space in order to visualise flight-condition transitions using the Generative Topographic Mapping (GTM) and a variant which supports simultaneous feature selection. We created an objective measure of the separation between different flight conditions in the visualisation space by calculating the Kullback-Leibler (KL) divergence between Gaussian mixture models (GMMs) fitted to each class: the higher the KL-divergence, the better the interclass separation. To find the optimal combination of sensors, they were considered in pairs, triples and groups of four sensors. The sensor triples provided the best result in terms of KL-divergence. We also found that the use of a variational training algorithm for the GMMs gave more reliable results.

AB - The main objective of the project is to enhance the already effective health-monitoring system (HUMS) for helicopters by analysing structural vibrations to recognise different flight conditions directly from sensor information. The goal of this paper is to develop a new method to select those sensors and frequency bands that are best for detecting changes in flight conditions. We projected frequency information to a 2-dimensional space in order to visualise flight-condition transitions using the Generative Topographic Mapping (GTM) and a variant which supports simultaneous feature selection. We created an objective measure of the separation between different flight conditions in the visualisation space by calculating the Kullback-Leibler (KL) divergence between Gaussian mixture models (GMMs) fitted to each class: the higher the KL-divergence, the better the interclass separation. To find the optimal combination of sensors, they were considered in pairs, triples and groups of four sensors. The sensor triples provided the best result in terms of KL-divergence. We also found that the use of a variational training algorithm for the GMMs gave more reliable results.

UR - http://www.scopus.com/inward/record.url?scp=84870666847&partnerID=8YFLogxK

U2 - 10.1109/MLSP.2012.6349808

DO - 10.1109/MLSP.2012.6349808

M3 - Conference contribution

SN - 978-1-4673-1024-6

T3 - Machine learning for signal processing

BT - IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2012

PB - IEEE

ER -

Gill WS, Nabney IT, Wells D. Helicopter vibration sensor selection using data visualisation. In IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2012. IEEE. 2012. (Machine learning for signal processing). https://doi.org/10.1109/MLSP.2012.6349808