Frequency-Dependent Peak-Over-Threshold algorithm for fault detection in the spectral domain

Aurélien Hazan (1) , Kurosh Madani (1)

1 SYNAPSE

An original novelty detection algorithm in the Fourier domain, using extreme value theory (EVT) is considered in this article. Periodograms may be considered as frequency-dependent random vari- ables, and this can be taken into account when designing statistical tests. Frequency-Dependent Peak-Over-Threshold (FDPOT) puts special em- phasis on the frequency dependence of extreme value statistics, thanks to Vector Generalized Additive Models (VGAM) estimation. An application is discussed in the field of mechanical vibrations. It is first shown that performance increases compared to POT detection. Then FDPOT is compared to state-of-the-art algorithms such as KPCA.

Mots clés: vibration; spectrum; detection; extreme; novelty

Domaines: Machine Learning [stat.ML]; Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]; Mécanique [physics.med-ph]

Aurélien Hazan, Kurosh Madani. Frequency-Dependent Peak-Over-Threshold algorithm for fault detection in the spectral domain. Proc. Of the 21th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN2013, Apr 2013, Bruges, Belgium. ⟨hal-00785382v2⟩