Gene expression programming-based approach for predicting the roller length of a hydraulic jump on a rough bed. (2nd November 2021)

• Record Type: Journal Article
• Authors: Azimi, Hamed; Bonakdari, Hossein; Ebtehaj, Isa
• Is Part Of: ISH journal of hydraulic engineering. Volume 27(2021)supplement 1
• Page Start: 77
• Page End: 87
• Publication Date: 2021-11-02
• Subjects: Hydraulic jump -- roller length -- rough bed -- partial derivative sensitivity analysis (PDSA) -- uncertainty analysis; Hydraulic engineering -- Periodicals; Hydraulic engineering -- India -- Periodicals; Hydraulic engineering; India; Periodicals; 627
• DOI: 10.1080/09715010.2019.1579058
• Languages: English
• ISSNs: 0971-5010

Abstract

ABSTRACT: Hydraulic jumps are used for dissipating water flow energy after ogee spillways, chlorinating wastewater flow and other hydraulic purposes. This paper presents a concise and practical relationship based on Gene Expression Programming (GEP) to estimate the roller length of a hydraulic jump on a rough bed. K-fold cross validation ( k = 10) is applied to validate the numerical results. Various models are developed in terms of the Froude number upstream of the hydraulic jump F 1, ratio of sequent depth h 2 / h 1 and the relative roughness k s / h 1 . The model results are analyzed and compared to determine the superior model. The MARE and RMSE obtained for the best model are 0.079 and 1.965, while the correlation coefficient ( R ), scatter index and BIAS are 0.959, 0.101 and 0.003, respectively. To evaluate each input variable's effect on the proposed model, Partial Derivative Sensitivity Analysis (PDSA) as well as uncertainty analysis is done for all GEP models. A number of artificial intelligence approaches are also compared. This study shows that the presented equation is simpler and more practical than other equations from prior studies.