Differential evolution algorithm in dynamic structural identification

In the present paper, a Differential Evolution algorithm is used to perform a structural identification of mass and stiffness properties of a frame structure. Identification is performed starting from exact data and, alternatively, from pseudo-experimental data, obtained by adding some statistic scattering to exact data. Different object functions are adopted and results are compared with those obtained adopting classical gradient method. Numerical results confirm that adopting both frequencies and modal shapes instead of frequencies only strongly increases the sensitivity of object function to identification parameters. Results obtained are then much less scattered, with coefficient of variation of the same order of magnitude of that of pseudo-experimental data.