Eventos Anais de eventos
COBEM 2017
24th ABCM International Congress of Mechanical Engineering
Fault Diagnostic for Rotating Machines using Bayesian Networks
Submission Author:
Natalia Tyminski , SP
Co-Authors:
Helio Fiori de Castro, Natalia Tyminski
Presenter: Natalia Tyminski
doi://10.26678/ABCM.COBEM2017.COB17-0809
Abstract
The maintenance based on the vibration signal is very common and used nowadays. The rotating machines, even when balanced and aligned, shows some level of vibration. This vibration can be very dangerous when reaches higher vibrations, the accident can be catastrophic. For this reason, it´s important the study of fault diagnostic, once that the failures of this machines contributes for the level of vibrations. This paper aims the use of Bayesian Network to do the fault diagnosis, whereas the symptoms generated by the failures are taken into account. The results showed are consistent, besides the prior probabilities considered are based on a few simulated results. The method proved to be consistent and efficiently.
Keywords
Fault Diagnosis, Bayesian Network, rotor
