Static and dynamic novelty detection methods for jet engine health monitoring.

Novelty detection requires models of normality to be learnt from training data known to be normal. The first model considered in this paper is a static model trained to detect novel events associated with changes in the vibration spectra recorded from a jet engine. We describe how the distribution o...

Mô tả đầy đủ

Chi tiết về thư mục
Những tác giả chính: Hayton, P, Utete, S, King, D, King, S, Anuzis, P, Tarassenko, L
Định dạng: Journal article
Ngôn ngữ:English
Được phát hành: 2007
_version_ 1826268512683819008
author Hayton, P
Utete, S
King, D
King, S
Anuzis, P
Tarassenko, L
author_facet Hayton, P
Utete, S
King, D
King, S
Anuzis, P
Tarassenko, L
author_sort Hayton, P
collection OXFORD
description Novelty detection requires models of normality to be learnt from training data known to be normal. The first model considered in this paper is a static model trained to detect novel events associated with changes in the vibration spectra recorded from a jet engine. We describe how the distribution of energy across the harmonics of a rotating shaft can be learnt by a support vector machine model of normality. The second model is a dynamic model partially learnt from data using an expectation-maximization-based method. This model uses a Kalman filter to fuse performance data in order to characterize normal engine behaviour. Deviations from normal operation are detected using the normalized innovations squared from the Kalman filter.
first_indexed 2024-03-06T21:10:49Z
format Journal article
id oxford-uuid:3e17d68d-7c26-4a35-ad9c-11d7e0760df1
institution University of Oxford
language English
last_indexed 2024-03-06T21:10:49Z
publishDate 2007
record_format dspace
spelling oxford-uuid:3e17d68d-7c26-4a35-ad9c-11d7e0760df12022-03-26T14:23:26ZStatic and dynamic novelty detection methods for jet engine health monitoring.Journal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:3e17d68d-7c26-4a35-ad9c-11d7e0760df1EnglishSymplectic Elements at Oxford2007Hayton, PUtete, SKing, DKing, SAnuzis, PTarassenko, LNovelty detection requires models of normality to be learnt from training data known to be normal. The first model considered in this paper is a static model trained to detect novel events associated with changes in the vibration spectra recorded from a jet engine. We describe how the distribution of energy across the harmonics of a rotating shaft can be learnt by a support vector machine model of normality. The second model is a dynamic model partially learnt from data using an expectation-maximization-based method. This model uses a Kalman filter to fuse performance data in order to characterize normal engine behaviour. Deviations from normal operation are detected using the normalized innovations squared from the Kalman filter.
spellingShingle Hayton, P
Utete, S
King, D
King, S
Anuzis, P
Tarassenko, L
Static and dynamic novelty detection methods for jet engine health monitoring.
title Static and dynamic novelty detection methods for jet engine health monitoring.
title_full Static and dynamic novelty detection methods for jet engine health monitoring.
title_fullStr Static and dynamic novelty detection methods for jet engine health monitoring.
title_full_unstemmed Static and dynamic novelty detection methods for jet engine health monitoring.
title_short Static and dynamic novelty detection methods for jet engine health monitoring.
title_sort static and dynamic novelty detection methods for jet engine health monitoring
work_keys_str_mv AT haytonp staticanddynamicnoveltydetectionmethodsforjetenginehealthmonitoring
AT utetes staticanddynamicnoveltydetectionmethodsforjetenginehealthmonitoring
AT kingd staticanddynamicnoveltydetectionmethodsforjetenginehealthmonitoring
AT kings staticanddynamicnoveltydetectionmethodsforjetenginehealthmonitoring
AT anuzisp staticanddynamicnoveltydetectionmethodsforjetenginehealthmonitoring
AT tarassenkol staticanddynamicnoveltydetectionmethodsforjetenginehealthmonitoring