6533b85dfe1ef96bd12be8b1

RESEARCH PRODUCT

Damage identification by Lévy ant colony optimization

Antonina PirrottaGiuseppe Fileccia ScimemiEleonora Riva SanseverinoGiulio Cottone

subject

business.industryComputer sciencedamage identification optimization levy acorAnt colony optimization algorithmsIdentification (biology)Pattern recognitionArtificial intelligenceSettore ICAR/08 - Scienza Delle Costruzionibusiness

description

This paper deals with the identification of incipient damage in structural elements by non-destructive test based on experimentally measured structural dynamical response. By applycation of the Hilbert transform to the recorded signal the so-called phase of the analytical signal is recovered and a proper functional is constructed in such a way that its global minimum gives a measure of the damage level, meant as stiffness reduction. Minimization is achieved by applying a modified Ant Colony Optimization (ACO) for continuous variables, inspired by the ants’ forageing behavior. The modification consists in the application of a new perturbation operator, based on alpha stable Lévy distribution. Numerical application to three degrees of freedom system is proposed to show the method and test the performance.

https://doi.org/10.1201/b10497-7