Inference of helicopter airframe condition

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

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

Abstract

The goal of this paper is to model normal airframe conditions for helicopters in order to detect changes. This is done by inferring the flying state using a selection of sensors and frequency bands that are best for discriminating between different states. We used non-linear state-space models (NLSSM) for modelling flight conditions based on short-time frequency analysis of the vibration data and embedded the models in a switching framework to detect transitions between states. We then created a density model (using a Gaussian mixture model) for the NLSSM innovations: this provides a model for normal operation. To validate our approach, we used data with added synthetic abnormalities which was detected as low-probability periods. The model of normality gave good indications of faults during the flight, in the form of low probabilities under the model, with high accuracy (>92 %).

Original languageEnglish
Title of host publication2013 IEEE International Workshop on Machine Learning for Signal Processing September 22-25, Southampton, United Kingdom. Proceedings MLSP2013
EditorsSaeid Sanei, Paris Smaragdis, Asoke Nandi, et al
PublisherIEEE
Number of pages6
ISBN (Print)978-1-4799-1180-6
DOIs
Publication statusPublished - 2013
Event16th IEEE international workshop on Machine Learning for Signal Processing - Southampton, United Kingdom
Duration: 22 Sep 201325 Sep 2013

Publication series

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

Workshop

Workshop16th IEEE international workshop on Machine Learning for Signal Processing
Abbreviated titleMLSP 2013
CountryUnited Kingdom
CitySouthampton
Period22/09/1325/09/13

Fingerprint

Airframes
Helicopters
Frequency bands
Innovation
Sensors

Keywords

  • condition monitoring
  • flight condition
  • non-linear model
  • signal processing
  • switching state space
  • vibration

Cite this

Gill, W. S., Nabney, I. T., & Wells, D. (2013). Inference of helicopter airframe condition. In S. Sanei, P. Smaragdis, A. Nandi, & et al (Eds.), 2013 IEEE International Workshop on Machine Learning for Signal Processing September 22-25, Southampton, United Kingdom. Proceedings MLSP2013 (Machine learning for signal processing). IEEE. https://doi.org/10.1109/MLSP.2013.6661960
Gill, Waljinder S. ; Nabney, Ian T. ; Wells, Daniel. / Inference of helicopter airframe condition. 2013 IEEE International Workshop on Machine Learning for Signal Processing September 22-25, Southampton, United Kingdom. Proceedings MLSP2013. editor / Saeid Sanei ; Paris Smaragdis ; Asoke Nandi ; et al. IEEE, 2013. (Machine learning for signal processing).
@inproceedings{94d9c92691af42a6ae0acc7d64c0c031,
title = "Inference of helicopter airframe condition",
abstract = "The goal of this paper is to model normal airframe conditions for helicopters in order to detect changes. This is done by inferring the flying state using a selection of sensors and frequency bands that are best for discriminating between different states. We used non-linear state-space models (NLSSM) for modelling flight conditions based on short-time frequency analysis of the vibration data and embedded the models in a switching framework to detect transitions between states. We then created a density model (using a Gaussian mixture model) for the NLSSM innovations: this provides a model for normal operation. To validate our approach, we used data with added synthetic abnormalities which was detected as low-probability periods. The model of normality gave good indications of faults during the flight, in the form of low probabilities under the model, with high accuracy (>92 {\%}).",
keywords = "condition monitoring, flight condition, non-linear model, signal processing, switching state space, vibration",
author = "Gill, {Waljinder S.} and Nabney, {Ian T.} and Daniel Wells",
year = "2013",
doi = "10.1109/MLSP.2013.6661960",
language = "English",
isbn = "978-1-4799-1180-6",
series = "Machine learning for signal processing",
publisher = "IEEE",
editor = "Saeid Sanei and Paris Smaragdis and Asoke Nandi and {et al}",
booktitle = "2013 IEEE International Workshop on Machine Learning for Signal Processing September 22-25, Southampton, United Kingdom. Proceedings MLSP2013",
address = "United States",

}

Gill, WS, Nabney, IT & Wells, D 2013, Inference of helicopter airframe condition. in S Sanei, P Smaragdis, A Nandi & et al (eds), 2013 IEEE International Workshop on Machine Learning for Signal Processing September 22-25, Southampton, United Kingdom. Proceedings MLSP2013. Machine learning for signal processing, IEEE, 16th IEEE international workshop on Machine Learning for Signal Processing, Southampton, United Kingdom, 22/09/13. https://doi.org/10.1109/MLSP.2013.6661960

Inference of helicopter airframe condition. / Gill, Waljinder S.; Nabney, Ian T.; Wells, Daniel.

2013 IEEE International Workshop on Machine Learning for Signal Processing September 22-25, Southampton, United Kingdom. Proceedings MLSP2013. ed. / Saeid Sanei; Paris Smaragdis; Asoke Nandi; et al. IEEE, 2013. (Machine learning for signal processing).

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

TY - GEN

T1 - Inference of helicopter airframe condition

AU - Gill, Waljinder S.

AU - Nabney, Ian T.

AU - Wells, Daniel

PY - 2013

Y1 - 2013

N2 - The goal of this paper is to model normal airframe conditions for helicopters in order to detect changes. This is done by inferring the flying state using a selection of sensors and frequency bands that are best for discriminating between different states. We used non-linear state-space models (NLSSM) for modelling flight conditions based on short-time frequency analysis of the vibration data and embedded the models in a switching framework to detect transitions between states. We then created a density model (using a Gaussian mixture model) for the NLSSM innovations: this provides a model for normal operation. To validate our approach, we used data with added synthetic abnormalities which was detected as low-probability periods. The model of normality gave good indications of faults during the flight, in the form of low probabilities under the model, with high accuracy (>92 %).

AB - The goal of this paper is to model normal airframe conditions for helicopters in order to detect changes. This is done by inferring the flying state using a selection of sensors and frequency bands that are best for discriminating between different states. We used non-linear state-space models (NLSSM) for modelling flight conditions based on short-time frequency analysis of the vibration data and embedded the models in a switching framework to detect transitions between states. We then created a density model (using a Gaussian mixture model) for the NLSSM innovations: this provides a model for normal operation. To validate our approach, we used data with added synthetic abnormalities which was detected as low-probability periods. The model of normality gave good indications of faults during the flight, in the form of low probabilities under the model, with high accuracy (>92 %).

KW - condition monitoring

KW - flight condition

KW - non-linear model

KW - signal processing

KW - switching state space

KW - vibration

UR - http://www.scopus.com/inward/record.url?scp=84893273885&partnerID=8YFLogxK

U2 - 10.1109/MLSP.2013.6661960

DO - 10.1109/MLSP.2013.6661960

M3 - Conference contribution

SN - 978-1-4799-1180-6

T3 - Machine learning for signal processing

BT - 2013 IEEE International Workshop on Machine Learning for Signal Processing September 22-25, Southampton, United Kingdom. Proceedings MLSP2013

A2 - Sanei, Saeid

A2 - Smaragdis, Paris

A2 - Nandi, Asoke

A2 - et al,

PB - IEEE

ER -

Gill WS, Nabney IT, Wells D. Inference of helicopter airframe condition. In Sanei S, Smaragdis P, Nandi A, et al, editors, 2013 IEEE International Workshop on Machine Learning for Signal Processing September 22-25, Southampton, United Kingdom. Proceedings MLSP2013. IEEE. 2013. (Machine learning for signal processing). https://doi.org/10.1109/MLSP.2013.6661960