Damage identification by a modified Ant Colony Optimization for not well spaced frequency systems

Abstract

Recently, it has been shown , that a damage detection strategy based on a proper functional calculated on the analytical signal of the structural dynamical response, consents to identify very low damage level. In this regard, they stressed the efficiency of Hilbert Transform to obtain the analytical response representation that shows more sensitivity for predicting damage with respect to the simple signal response. Then, a damage identification procedure based on the minimization of the difference between theoretical and measured data was proposed with satisfactory results. Unfortunately, this procedure, since the need of use of band pass filter around the natural frequency of the system, fails for structures having closed natural frequencies. By the way, performing procedures for sharply detecting damage in not well spaced frequency structure is a hoary problem. Aim of this paper is to extend the aforementioned procedure to these systems. To aim at this, it is desirable to go further insight into optimization algorithms, suitable for this kind of systems. For instance, it will be interesting considering, the ant colony optimization algorithm (ACO) that is a probabilistic technique for solving computational problems which can be reduced to finding good paths through graphs. Recently ACO has been extended to continuous domain and labeled as ACOð. A novel aspect of the proposed paper is introducing ACOð into the previous procedure avoiding the use of filters, such that may be available for not well spaced frequency system. Moreover, it will be desirable avoiding the use of Hilbert transform, that means apply the identification procedure directly on the acceleration responses and not on the analytical signal response. Therefore, in this paper it will be introduced a procedure for detecting damage in structures having close frequencies, without using analytical signal response.