Affiliation:
1. University of Memphis, USA
Abstract
Scour is a critical issue that impacts the safety and strength of bridges. Precise scour forecasts around bridge piers can provide useful data for bridge engineers to bring preventive actions. This study uses long short-term memory (LSTM) neural network with Bayesian optimization to forecast the scour around the bridges and piers. The LSTM network was trained and tested using only scour depth data from a calibrated numerical model. The outcomes indicate that the proposed LSTM model provides precise scour depth forecasts. The study presents the performance of the LSTM model for predicting scour depth around bridge piers, which can help enhance the safety and stability of bridges. The model has shown acceptable outcomes, with a rank correlation equal to 0.9866 in the training stage and 0.9655 in the testing stage. Moreover, the LSTM model was used to forecast the scour depth for 11 minutes.
Reference64 articles.
1. Computational fluid dynamics modeling of abutment scour under steady current using the level set method
2. Modeling of Local Scour Depth Around Bridge Pier Using FLOW 3D
3. Ali, A. S. A.. (2021). Republic of Turkey Gaziantep University Graduate School of Natural and Applied Sciences.
4. Machine Learning Model for Estimation of Local Scour Depth around Cylindrical Bridge Piers
5. Ali, A. S., Jazaei, F., Ashiq, M. M., Bakhshaee, A., & Alasta, M. S. (2022, December). Numerical Model For Simulation Of Scour Around Bridge Abutment. In Fall Meeting 2022. AGU.