<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>Local scour around bridge piers impairs the stability of bridges’ structures. Therefore, a delicate estimation of the local scour depth is vital in designing the bridge piers foundations. In this research, MATLAB software was used to train artificial neural network (ANN) models with four hundred laboratory datasets from different laboratory studies, including five parameters: pier diameter, flow depth flow velocity, critical sediment velocity, sediment particle size, and equilibrium local scour depth. The outcomes present that the ANN model with the Levenberg-Marquardt algorithm and 11 nodes in the single hidden layer gives an accurate estimation better than other ANN models trained with different training algorithms based on the regression results and mean squared error values. Besides, the ANN model accurately provides predicted local scour depth and is better than linear and nonlinear regression models. Furthermore, sensitivity analysis shows that removing pier diameter from training parameters diminishes the reliability of prediction.</p>
</abstract>

Local scour around bridge piers impairs the stability of bridges’ structures. Therefore, a delicate estimation of the local scour depth is vital in designing the bridge piers foundations. In this research, MATLAB software was used to train artificial neural network (ANN) models with four hundred laboratory datasets from different laboratory studies, including five parameters: pier diameter, flow depth flow velocity, critical sediment velocity, sediment particle size, and equilibrium local scour depth. The outcomes present that the ANN model with the Levenberg-Marquardt algorithm and 11 nodes in the single hidden layer gives an accurate estimation better than other ANN models trained with different training algorithms based on the regression results and mean squared error values. Besides, the ANN model accurately provides predicted local scour depth and is better than linear and nonlinear regression models. Furthermore, sensitivity analysis shows that removing pier diameter from training parameters diminishes the reliability of prediction.

Artificial Neural Network for Estimation of Local Scour Depth Around Bridge Piers

Faculty of Engineering, Department of Civil Engineering

Archives of Hydro-Engineering and Environmental Mechanics

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Local scour around bridge piers impairs the stability of bridges’ structures. Therefore, a delicate estimation of the local scour depth is vital in designing...

Artificial Neural Network for Estimation of Local Scour Depth Around Bridge Piers

Online veröffentlicht: 18. Jan. 2022

Seitenbereich: 87 - 101

Eingereicht: 11. Jan. 2021

Akzeptiert: 22. Aug. 2021

DOI: https://doi.org/10.2478/heem-2021-0005

Schlüsselwörter
artificial neural network, bridge pier, hydraulics, local scour

© 2021 Ahmed Shakir Ali Ali et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Artificial Neural Network for Estimation of Local Scour Depth Around Bridge Piers

Online veröffentlicht: 18. Jan. 2022

Seitenbereich: 87 - 101

Eingereicht: 11. Jan. 2021

Akzeptiert: 22. Aug. 2021

DOI: https://doi.org/10.2478/heem-2021-0005

Schlüsselwörterartificial neural network, bridge pier, hydraulics, local scour

© 2021 Ahmed Shakir Ali Ali et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Schlüsselwörter
artificial neural network, bridge pier, hydraulics, local scour