Author(s): Reza Mohammadpour, Aminuddin Ab. Ghani

Linked Author(s): Reza Mohammadpour Ghalati

Keywords: Neural network, GEP, local scour, complex piers scour, erosion

Abstract:

Accurate estimation of scour depth at complex pier leads to an economic and correct design for level of foundation. Due to technical and economic reasons, complex piers (pier including foundation) are more general for bridge design. In this study, two robust techniques, artificial neural networks (ANNs) and Gene Expression Programming (GEP) were employed for prediction of scour depth at complex piers. A wide range of dataset was collected from present study as well as literature, the clear water condition was chosen for experimental tests. The result shows that the scour depth at complex piers is a function of pier diameter (bc) and foundation level (Y). The RBF (Radial Base Function) network with R2 =0.945 and RMSE =0.031 provides better prediction in comparison with previous equations and the GEP techniques (R2 =0.811and RMSE=0.263). A formula was developed using GEP to predict local scour at complex piers. Although, accuracy of the RBF is higher than GEP, the GEP based formula is more useful for practical purposes and can be easily employed to predict the depth of scour at complex piers. This paper highlights that the mentioned techniques can be successfully used for accurate estimation of local scour and lead to the protection of bridge piers. (2642, 72, 316)

DOI:

Year: 2017