In a structural health monitoring framework, most of the diagnostic and prognostic tasks require the solution of inverse problems, such as the identification of damage parameters (e.g., damage position and its extent) or the estimation of damage evolution model parameters from the observation of damage dependent features. Bayesian approaches have proven to be successful for inverse problem solution, by the evaluation of the posterior distribution of a vector of parameters of interest, conditioned on the observations of some signal features, which relies on the direct calculation of the observations likelihood. Unfortunately, for realistic structures, a numerical simulation might be required for the evaluation of each sample likelihood, which can make the whole procedure for the posterior pdf estimation computationally unfeasible. In this work, this problem is solved by leveraging on surrogate modelling. Sequential importance Monte-Carlo sampling, also known as particle filter, is used as a general framework for the health state estimation and prognosis of a skin panel subject to fatigue crack growth, while observing the strain field pattern acquired at some specific locations. A surrogate model consisting of an artificial neural network, trained on a set of analytical simulations, is used to predict the strain as a function of the crack position and length, thus allowing a fast calculation of the strain observation likelihood. The algorithm is tested with reference to an analytic case study of a crack propagating in an infinite plate, allowing for a simultaneous diagnosis of the crack position and length, as well as a realtime updating of the evolution model parameters and system prognosis.

Surrogate modelling for observation likelihood calculation in a particle filter framework for automated diagnosis and prognosis

C. Sbarufatti;F. Cadini;M. Giglio
2018-01-01

Abstract

In a structural health monitoring framework, most of the diagnostic and prognostic tasks require the solution of inverse problems, such as the identification of damage parameters (e.g., damage position and its extent) or the estimation of damage evolution model parameters from the observation of damage dependent features. Bayesian approaches have proven to be successful for inverse problem solution, by the evaluation of the posterior distribution of a vector of parameters of interest, conditioned on the observations of some signal features, which relies on the direct calculation of the observations likelihood. Unfortunately, for realistic structures, a numerical simulation might be required for the evaluation of each sample likelihood, which can make the whole procedure for the posterior pdf estimation computationally unfeasible. In this work, this problem is solved by leveraging on surrogate modelling. Sequential importance Monte-Carlo sampling, also known as particle filter, is used as a general framework for the health state estimation and prognosis of a skin panel subject to fatigue crack growth, while observing the strain field pattern acquired at some specific locations. A surrogate model consisting of an artificial neural network, trained on a set of analytical simulations, is used to predict the strain as a function of the crack position and length, thus allowing a fast calculation of the strain observation likelihood. The algorithm is tested with reference to an analytic case study of a crack propagating in an infinite plate, allowing for a simultaneous diagnosis of the crack position and length, as well as a realtime updating of the evolution model parameters and system prognosis.
2018
EWSHM 2018 9th European Workshop on Structural Health Monitoring (EWSHM 2018)
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/1073508
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? ND
social impact