Structural Health Monitoring of Underground Metro Tunnel by Identifying Damage Using ANN Deep Learning Auto-Encoder

Due to the complexity of underground environmental conditions and operational incidents, advanced accurate monitoring metro shield tunnel structures is crucial for maintenance prevention mishaps. In past few decades, numerous deep learning-based damage identification studies have been conducted on aboveground civil infrastructure. However, a tunnels. This paper presents study Based previous experimental studies, numerical model was utilized, vibration data obtained from under moving load analysis used evaluation. An existing auto-encoder (DAE) that can support neural networks utilized detect structural accurately by incorporating raw signals. The dynamic FEM with different severity levels at locations elements structure. addition, root mean square (RMS) locate in model. results were compared schemes white noise, varying damage, an intact state. To test applicability proposed framework small dataset, approach also investigate simply supported beam two methods (SVM LSTM). show DAE-based feasible efficient identification, size evaluation, localization comparison models.