A Comparison of Two Different Approaches for the Damage Idemtification Problem

This work deals with a structural damage detection problem using displacement measurements as experimental data to be used by two different methodologies: the conjugate gradient method with the adjoint equation and an artificial neural network. Both techniques have been employed in order to place and quantify the time-variable damage in a simple truss structure. Numerical experiments have been carried out with synthetic experimental data considering a noise level of 1%. Good recoveries have been achieved with both methodologies and estimation errors have been reported in order to make their comparison possible.