Structural damage recognition based on wavelet transform and improved most valuable player algorithm

Abstract

The probability of damage to infrastructure increases with age. Timely health monitoring of infrastructure is essential since it is linked to the safety of people's lives and property. The study employed the wavelet transform approach to discretize the wavelet transform of the observed vibration patterns of the damaged structure in order to identify and localize the structural damage in infrastructure. The study also employed natural excitation techniques to obtain the structural multi-order modal parameters. To quantify the degree of structural damage, the study designed an objective function for damage quantification and improved the most valuable player algorithm. The study avoided the most valuable player algorithm from falling into local optimality by introducing the elite inverse strategy and simplex strategy. The wavelet transform and the improved most valuable player algorithm were able to successfully identify and localize the structural damages and quantify the degree of the damages, thereby providing technical support for the daily monitoring of the infrastructure.