Fast Bayesian updating of large-scale finite element models using CMS technique and surrogate models

A Bayesian probabilistic framework for parameter estimation is applied for updating large-order finite element models of structures using response measurements. Fast and accurate component mode synthesis (CMS) techniques are proposed, consistent with the finite element model parameterization, to achieve drastic reductions in computational effort. Further computational savings are achieved by adopting heuristic approximations based on surrogate models. The computational efficiency and accuracy of the proposed techniques is demonstrated by updating a finite element model of a bridge involving hundreds of thousands of degrees of freedom.