Document Detail

Static and dynamic novelty detection methods for jet engine health monitoring.
MedLine Citation:
PMID:  17255049     Owner:  NLM     Status:  MEDLINE    
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.
Paul Hayton; Simukai Utete; Dennis King; Steve King; Paul Anuzis; Lionel Tarassenko
Related Documents :
24022319 - Comparison of methods for handling missing covariate data.
23005279 - Evidence for an excess of b[over ¯]→d^{(*)}τ^{-}ν[over ¯]_{τ} decays.
9691059 - A bayesian characterization of hardy-weinberg disequilibrium.
23831259 - The association between tumor necrosis factor alpha promoter polymorphisms and ankylosi...
23763419 - Televised antismoking advertising: effects of level and duration of exposure.
24269189 - Characterization of α-cypermethrin exposure in egyptian agricultural workers.
18547859 - Human heart beat analysis using a modified algorithm of detrended fluctuation analysis ...
1938139 - Determination of the number and waveshapes of event related potential components using ...
23999189 - Imodfit: efficient and robust flexible fitting based on vibrational analysis in interna...
Publication Detail:
Type:  Journal Article; Review    
Journal Detail:
Title:  Philosophical transactions. Series A, Mathematical, physical, and engineering sciences     Volume:  365     ISSN:  1364-503X     ISO Abbreviation:  Philos Trans A Math Phys Eng Sci     Publication Date:  2007 Feb 
Date Detail:
Created Date:  2007-01-26     Completed Date:  2007-03-27     Revised Date:  2013-04-24    
Medline Journal Info:
Nlm Unique ID:  101133385     Medline TA:  Philos Trans A Math Phys Eng Sci     Country:  England    
Other Details:
Languages:  eng     Pagination:  493-514     Citation Subset:  IM    
Department of Engineering Science, University of Oxford, Oxford OX1 3PJ, UK.
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Aircraft / instrumentation*
Computer Simulation
Construction Materials / analysis*
Engineering / instrumentation,  methods
Equipment Design
Equipment Failure Analysis / instrumentation,  methods*
Maintenance / methods
Materials Testing / methods*
Models, Theoretical*
Signal Processing, Computer-Assisted

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine

Previous Document:  An automated damage identification technique based on vibration and wave propagation data.
Next Document:  The application of machine learning to structural health monitoring.