CWT-CNN-Based Damage Localization in plate like structure

This study proposes a novel hybrid approach integrating Continuous Wavelet Transform (CWT) and Convolutional Neural Networks (CNN) for automated structural damage detection and localization in steel plates. A numerical finite element model of a square steel plate with fixed boundary conditions is developed to obtain the vibration characteristics and mode shapes under intact and damaged conditions. The differences between the first mode shapes of the intact and damaged plates are processed using CWT to extract localized damage-sensitive features at multiple scales. The resulting wavelet representations are converted into images and used as input to a CNN model for automated identification and localization of damaged regions. The results demonstrate that the proposed CWT–CNN framework can effectively distinguish damage locations and provides reliable prediction performance under the considered numerical conditions. The proposed methodology offers a promising and intelligent framework for structural health monitoring, with future scope for validation using experimental and real-time vibration data.