Damage Detection in Steel Structures Using Wavelet Transformation and Generalized Genetic Algorithm

Abstract

A new two-stage damage identification process determining damage severity and location in elements is put forward in this paper. In this approach, wavelet analysis is implemented to extract features sensitive to damage from cross-model modal strain energy and cross model cross mode euqation solution due to its strength of denoising. Meanwhile, an improved differential generalized algorithm is constructed to analyze the extent of damage. Non-dominated sorting techinique and mode combination sensitivity analysis are combined to handle constraints in optimization problem. The numerical results prove that the proposed methodology performs more accurately in damage location and quantificaiton than traditional damage recognition approaches.