1- Tarbiat Modares University
Abstract: (5950 Views)
Many empirical methods for estimating LSTR have been introduced by scientists during the recent decades, but these methods have been calibrated and applied under limited conditions of bed profile and specific range of bed sediment size. The existing empirical relations are linear or exponential regressions based on the observation and measurements data and there’s a great potential to build more accurate models to predict sediment transport phenomena by means of Soft computation approach. Contemporarily soft computing (SF) models, Artificial Neural Networks (ANN), Adaptive Neuro-Fuzzy Inference System (ANFIS) have been accepted as potentially valuable tool for modeling and forecasting complex nonlinear systems. In other words, SF is very helpful when the mathematical and physical scheme couldn’t propose accurate solution to the encountered problem. The main advantage of SF model is that the accurate detail of the problems is not required. A comprehensive comparison between both ANN and ANFIS models and the existing empirical formulae will be presented to demonstrate capacity of ANNs.
Type of Study:
Research Paper |
Subject:
Environmental Study Received: 2014/02/17 | Accepted: 2014/08/4