Optimized ANFIS-Genetic Algorithm-Particle Swarm Optimization Model for Estimation of Side Orifices Discharge Coefficient

Generally, side orifices are installed on side walls of the main channels for regulating and controlling water. In this study, a novel hybrid model was developed in order to estimate discharge coefficient of circular and rectangular side orifices. The model was obtained using combination of ANFIS, Genetic Algorithm and Particle Swarm Optimization. Then, 6 different models were defined for each ANFIS and hybrid models. Also, Monte Carlo Simulations (MCs) were used to survey abilities of the numerical models. Additionally, k-fold cross validation (k=5) was employed to validate the numerical models. Next, using sensitivity analysis, the superior model and the most effective parameters were identified. The best model simulated the discharge coefficient using all input variables this with high accuracy. For instance, the correlation coefficient and scatter index for the model were computed 0.856 and 0.027, respectively. Furthermore, the dimensionless parameter “the ratio of height of the side orifice to the side orifice diameter (W/D) was identified as the most effective variable input.