Reservoir Dam Inflow Modeling using Wavelet-Base Neural Network and Multi Layer Perceptron Network

Reservoir dam management needs the high accurate model to prediction of basin balance. Intelligence models could present more accurate estimate data for patterns simulation. But potential of learning process for each system has some restrictions. Conjunction of mother wavelets function and multi layer perceptron networks create wavelet neural networks (WNN) that recently attracted great interest, because of their advantages over MLP network mother wavelet function applied, the hybrid wavelet neural network (WNNs) employing nonlinear wavelet basis, which are localized in both the time space and frequency space, has been developed. Wavelet neural network applied for monthly inflow of Nazloochaei Reservoir Dam. The result of MLP model shows the 86% in training and 79% in model testing. But result of WNN shows that RASP Wavelet Neural Network in the best topology present 90% accuracy in training phase and 83% in testing of model. Comparison WNN and MLP show that the RASP optimized daughter wavelet could have better accuracy in monthly inflow of Nazloochaei Reservoir Dam modeling. RASP Mother Wavelet, Wavelet Neural Networks, daughter wavelet, Reservoir dam management.