Estimation of Discharge Coefficient of Triangular Side Orifices using Generalized Structure Group Method of Data Handling

In the study, the discharge coefficient of triangular side orifices is modeled by a new evolutionary algorithm entitled "generalized structure group method of data handling (GSGMDH)". Initially, the variables affecting the discharge coefficient are discerned and six different GSGMDH models are defined. After that, the experimental data are divided into two sub-groups, meaning that 70% of the experimental measurements are used for training and 30% for testing. Then, the superior model is introduced through the analysis of all GSGMDH models. This model estimates the discharge coefficient values with remarkable accuracy. For example, in the testing mode of the superior model, the correlation coefficient (R), Scatter Index (SI) and the Nash-Sutcliff proficiency coefficient (NSC) are calculated 0.999, 0.0008 and 0.999, respectively. Furthermore, the sensitivity analysis of the input parameters reveals that the Froude number (F1) and the ratio of the main channel width to the side orifice length (B/L) are the most effective input parameters. After that, the results yielded by the GSGMDH superior model are compared with the GMDH results to prove that the accuracy of the GSGMDH model is higher