Clustering for unsupervised fault diagnosis in nuclear turbine shut-down transients - Archive ouverte HAL Access content directly
Journal Articles Mechanical Systems and Signal Processing Year : 2015

Clustering for unsupervised fault diagnosis in nuclear turbine shut-down transients

Abstract

Empirical methods for fault diagnosis usually entail a process of supervised training based on a set of examples of signal evolutions " labeled " with the corresponding, known classes of fault. However, in practice, the signals collected during plant operation may be, very often, " unlabeled " , i.e., the information on the corresponding type of occurred fault is not available. To cope with this practical situation, in this paper we develop a methodology for the identification of transient signals showing similar characteristics, under the conjecture that operational/faulty transient conditions of the same type lead to similar behavior in the measured signals evolution. The methodology is founded on a feature extraction procedure, which feeds a spectral clustering technique, embedding the unsupervised Fuzzy C-Means (FCM) algorithm, that evaluates the functional similarity among the different operational/faulty transients. A procedure for validating the plausibility of the obtained clusters is also propounded based on physical considerations. The methodology is applied to a real industrial case, on the basis of 148 shutdown transients of a Nuclear Power Plant (NPP) steam turbine.
Fichier principal
Vignette du fichier
15_Clustering for Unsupervised Fault Diagnosis in Nuclear Turbine Shutdown Transients.pdf (1.55 Mo) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-01265643 , version 1 (01-02-2016)

Identifiers

Cite

Piero Baraldi, Francesco Di Maio, Marco Rigamonti, Enrico Zio, Redouane Seraoui. Clustering for unsupervised fault diagnosis in nuclear turbine shut-down transients. Mechanical Systems and Signal Processing, 2015, 58-59, pp.160-178. ⟨10.1016/j.ymssp.2014.12.018⟩. ⟨hal-01265643⟩
437 View
209 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More