Modeling Discharge Coefficient of Side Weir on Converging Channel Using Extreme Learning Machine

In this study, the discharge coefficient of side weirs located on converging channels was simulated for the first time using a new method of Extreme Learning Machine (ELM). To examine the accuracy of the numerical model, the Monte Carlo simulations were used and the experimental values validation was conducted by the k-fold cross validation method. Then, the input parameters were detected for simulating the discharge coefficient. Subsequently, the number of the Extreme Learning Machine hidden layer neurons was determined using a trial and error process. In the next step, the most optimized activation function was also obtained. Then, using the input parameters, six ELM models were developed and the superior model and the most effective input parameter were identified through a sensitivity analysis. The superior model estimated the discharge coefficient values with an acceptable accuracy. For example, the values of the indices R2 and MAPE for this model were estimated 0.963 and 5.135, respectively and the Froude number at the downstream of the side weir (Fd) was introduced as the most effective parameter. Then, a relationship was provided for the superior model.