Generative Adversarial Networks for Labelled Vibration Data Generation

As Structural Health Monitoring (SHM) being implemented more over the years, use of operational modal analysis civil structures has become significant for assessment and evaluation engineering structures. Machine Learning (ML) Deep (DL) algorithms have been in structural damage diagnostics last couple decades. While collecting vibration data from is a challenging expensive task both undamaged damaged cases, this paper, authors are introducing Generative Adversarial Networks (GAN) that built on Convolutional Neural Network (DCNN) using Wasserstein Distance generating artificial labelled to be used diagnostic purposes. The named developed model “1D W-DCGAN” successfully generated which very similar input. methodology presented paper will pave way generation numerous future applications SHM domain.