Abstract
Local scour is considered a natural phenomenon resulting from the erosive effect of flowing water. In this study, gene expression programming (GEP) was applied to predict the local scour depth around bridge piers for both live-bed and clear-water conditions. Input variables that were considered as an effective parameter on scour depth modeling include properties of sediment size, pier geometry, and upstream flow conditions. The training and testing stages of GEP models were carried out using laboratory data sets collected from different literature. Furthermore, these data sets were used to develop nonlinear regression equations to predict scour depth for both flow conditions. Testing results indicated that the GEP models predicted the scour depth with lower error and higher accuracy than those performed using new-regression models and previous empirical equations (RMSE = 0.09, MAE = 0.07, and R2 = 0.9) for live-bed GEP model and (RMSE = 0.12, MAE = 0.08, and R2 = 0.92) for clear-water GEP model.
Disclosure statement
No potential conflict of interest was reported by the authors.
Additional information
Notes on contributors
Layla Ali Mohammed Saleh
Layla Ali Mohammed Saleh is currently a lecturer at the Department of Civil Engineering, Kerbala University, Iraq. In 2011, she received a Master's degree in Hydraulic Structure Engineering from Kufa University, Iraq. Areas of interest include bridge scouring, seepage, water quality, and soft computing.
Sumayah Amal Al-din Majeed
Sumayah Amal Al-din Majeed is currently a lecturer at the Department of Civil Engineering, Kerbala University, Iraq. In 2014, she received a Master's degree in Water Resources Engineering from Babylon University, Iraq.
Fatin Abd el-kadhium M. Alnasrawi
Fatin Abd el-kadhium M. Alnasrawi is currently a lecturer at the Department of Civil Engineering, Kerbala University, Iraq. In 2011, she received a Master's degree in Environmental Engineering from Babylon University, Iraq.