Nondestructive damage identification in beam-like structures via mode shape-based wavelet analysis method and artificial neural networks

Abstract

The development of non-destructive methods for structural health monitoring is essential. In recent years, research has required interdisciplinary and cross-disciplinary integration such as civil engineering, digital signal processing, and artificial intelligence. This integration aims to improve the efficiency and accuracy of the structural health monitoring method. From the research motivation, this paper proposes a nondestructive damage identification in beam-like structures using mode shape-based Wavelet analysis method and artificial neural networks (ANNs). Firstly, the theory of the Wavelet analysis method and the ANNs is described. Then, a finite element model of a simply supported beam is simulated. The investigated damage scenarios include cases of one, two, and three damage locations with different damage levels on the beam. The reliability of the simulation results is validated by comparing the numerical natural frequencies with theoretical results. Finally, the occurrence, location, and level of damage on the beam are accurately identified by using the proposed method. The results of this study demonstrate that the mode shape-based Wavelet analysis method, when combined with ANNs, achieves high effectiveness for damage identification in beam-like structures.