Structural Health Monitoring with Deep Learning

Electromechanical impedance (EMI) Method is a popular Structural Health Monitoring (SHM) techniques for monitoring the integrity of a mechanical structure. The EMI method is highly sensitivity to small damage. However, it also has a well-known issue, an impedance signal can be changed by other ambient variations. It has the difficulty in damage measurement with the index-based measurement methods, such as RMSD (Root Mean Square Deviation). In this article, we studied the application of the Deep Learning technique to address this issue. An experimental setup was designed for applying the EMI method to monitor the integrity of a metallic structure. The damage classification process has been carried out with a Deep learning tool. This preliminary study demonstrated a very positive result with a reliable measurement with the testing configuration.