Eventos Anais de eventos
COBEM 2019
25th International Congress of Mechanical Engineering
Identification of the fault parameters in a rotor system by Bayesian inference with polynomial chaos expansion
Submission Author:
Helio Fiori de Castro , SP , Brazil
Co-Authors:
Gabriel Garoli, Diogo Stuani Alves, Felipe Wenzel da Silva Tuckmantel, Katia Lucchesi Cavalca Dedini, Helio Fiori de Castro
Presenter: Helio Fiori de Castro
doi://10.26678/ABCM.COBEM2019.COB2019-0467
Abstract
Rotating machines are vastly used in industry, being commonly subject to faults that should be correctly identified for right maintenance purpose. Several methods to identify fault parameters are available, however, when the rotor system have more than one fault the identification process becomes more complex. Therefore, it is proposed a stochastic approach based on Bayesian Inference for the identification of fault parameters. The solution of the inference is commonly made by a Monte Carlo via Markov Chains methods, nevertheless it is very time consuming. The generalized Polynomial Chaos Expansion with the Stochastic Collocation can solve the inference in a shorter time. The method approximates part of the inference by a polynomial series, resuming the problem into evaluate the expansion coefficients, which is done by the Stochastic Collocation. Thus, the Bayesian Inference is used to identify the parameters of mass unbalance and angular misalignment introduced to the theoretical rotor system. The method shows satisfactory results evaluating the fault parameters when the faults are considered separately.
Keywords
generalized polynomial chaos expansion, unbalance, Misaligngment, Rotordynamics, Bayesian inference
