Estimating Free and Submerged Hydraulic Jump's Length in Horizontal and Slopping Channels Using Support Vector Regression

Hydraulic jump is the most common method for kinetic energy dissipating at downstream of spillways, chutes and gates. Several relations have been proposed to estimate the length of hydraulic jump, but the results of these equations are not general and acceptable due to the uncertainty of the functions. Consequently, it is essential to estimate the hydraulic jump length, accurately. In this paper, hydraulic jump length was estimated for free and submerged hydraulic jumps on horizontal and slopping smooth beds using support vector regression as one of the machine learning methods and the rate of influence of input parameters in each jump was analyzed. Totally, 294 patterns of the observed data were used for training and testing processes of the four kinds of hydraulic jump models. Comparison between support vector regression (SVR), classical and empirical equations and gene expression programming (GEP) method showed the noticeable efficiency of the support vector regression.