Eventos Anais de eventos
COBEM 2019
25th International Congress of Mechanical Engineering
STRUCTURAL DAMAGE IDENTIFICATION IN BEAMS WITH A NEW ADAPTIVE POPULATION-BASED MARKOV CHAIN MONTE CARLO METHOD
Submission Author:
Josiele da Silva Teixeira , RJ , Brazil
Co-Authors:
Josiele da Silva Teixeira, Leonardo Stutz, Antônio Silva Neto, Diego Knupp
Presenter: Josiele da Silva Teixeira
doi://10.26678/ABCM.COBEM2019.COB2019-1265
Abstract
The formulation and solution of the inverse problem of structural damage identification is addressed in the present work, built on the Bayesian approach. A cohesion field, spatially discretized by the finite element method, is considered in order to continuously describes the damage state of the structure. The inverse problem of damage identification is formulated from the Bayesian point of view, whose objective is to estimate the posterior probability densities of the nodal cohesion parameters. The standard Markov Chain Monte Carlo method (MCMC) and a proposed adaptive population-based MCMC method are considered in order to approximate the posterior probability distributions of the parameters of interest. In this approach, prior information on the sought parameters can be used and the uncertainties concerning the measured data can be quantified in the estimation of the cohesion parameters. A simply supported Euler–Bernoulli beam is considered to assess, by means of numerical simulations, the proposed approach.
Keywords
Inverse problems, Damage Identification, Bayesian inference, Markov Chain Monte Carlo Methods
