Prediction of Hydraulic Jump Characteristics on Rough Bed Using Artificial Neural Network and Genetic Programming

In the present research the hydraulic jump characteristics such as depth and length of the jump on rough beds as functions of roughness height and initial Froude number were simulated using artificial neural network (ANN) and genetic programming (GP) models. In the experiments, initial Froude numbers and roughness ratios were in the range of 1.9 to10 and 0.085 to 2.025, respectively.
Totally, 454 sets of the observed data were used in training and testing process of the proposed ANN and GP models. The results of the both artificial neural network and genetic programming models had good agreements with the measured data. Also the results of these models were compared with the known empirical equations for rough beds. It was shown that the ANN and GP models had less computational errors than the empirical equations. Also, the outputs of the GP model were presented in the form of mathematical equations and tree graph.