Finite Element Updating using Cross-Entropy Combined with Random-Field Theory

In this paper, the possibility of introducing random field theory into the cross-entropy algorithm is studied. Cross-entropy algorithm is an optimization process that Walsh and González (2009) use to estimate the stiffness distribution of a structure given a set of displacements. Although this method has been successfully tested, many lines of improvement are still opened. Random field theory is incorporated into the algorithm in an attempt to account for spatial variability of stiffness throughout the structure. For this purpose, a correlation function, variable in space, is defined and, as a result, a modification of the algorithm is proposed. The modified algorithm is then tested using numerical simulations in a scenario consisting of a simply supported beam.