PREDICTING THE INPUT FLOW INTO THE DAM RESERVOIR USING THE NEURAL NETWORK (CASE STUDY: DEZ DAM), IRANAmir Pourhaghi, Ali Mohammad Akhond Ali, Feridon Radmanesh, Hassan Torabi Podeh and Abozar Solgi
Awareness of the input flow into the dams reservoirs in future time periods is of the most important and valuable information which contributes the planners policy making in managing and dedicating the water resources. This research has been performed to model the amount of input flow into the Dezdams reservoir using the neural network models. To model by neural network, the monthly discharge data have been considered as the input data and the data before the model execution has been considered as the output data of the network. After examining different neural networksfitness the appropriate models to predict the flow was selected and at the end using the integrated genetic algorithm the number of appropriate layers and neurons in each layer and the best repetition number were specified. Finally, the results showed that the neural network model of GFF with the tangent hyperbolic tangent transfer function and Conjugate gradient training rule had a better efficiency than other models.
Enter your contact information below to receive full paper.