Comparative study of soft computing models for prediction of scour below two symmetric crossing jets

• Published in: Arabian journal of geosciences Vol. 15; no. 21
• Main Authors: Naini, Shayan, Karami, Hojat, Hosseini, Khosrow
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.11.2022; Springer Nature B.V
• Subjects: Adaptive systems; Artificial neural networks; Back propagation; Back propagation networks; Comparative analysis; Comparative studies; Computational neuroscience; Correlation coefficient; Correlation coefficients; Earth and Environmental Science; Earth science; Earth Sciences; Empirical analysis; Fuzzy logic; Hydraulic structures; Hydraulics; Machine learning; Methods; Neural networks; Original Paper; Radial basis function; Regression analysis; Regression models; Robustness (mathematics); Scour; Scouring; Sensitivity analysis; Soft computing; Support vector machines; Regression equations; Scour below crossing jets; Machine learning methods
• ISSN: 1866-7511; 1866-7538
• DOI: 10.1007/s12517-022-10910-z

The scour estimation downstream of the dam’s hydraulic structures is a serious issue and has long been considered an important topic by hydraulic engineers. The literature shows that the computational methods are good alternatives for predicting the scour in hydraulic structures in which the conventional methods demonstrate shortcomings. In the present paper, various machine learning techniques have been used for the first time to predict the scour below the two symmetric crossing jets. Four types of artificial neural networks (ANNs), including feedforward back-propagation (FFBP), cascade-forward back-propagation (CFBP), radial basis function (RBF), generalized regression neural network (GRNN) along with the adaptive neuro-fuzzy inference system (ANFIS), and support vector regression (SVR) were employed and evaluated. Two important scour hole characteristics, namely the maximum scour hole depth and its location relative to the scour hole origin, were predicted at three crossing angles. The soft computing models were also compared with the traditional empirical methods. The results indicated that the applied computing techniques perform satisfactorily and improve the results remarkably. They can estimate the scour more precisely than the regression models and can be considered robust alternative tools. A detailed sensitivity analysis was also performed on the FFBP that shows the crossing angle has a powerful effect on the predictions. The SVR improved the results to 47% and 31.71% for the scour hole depth and its location, respectively, in terms of the correlation coefficient.