Title: Evaluation of fluvial flow effects on tidal characteristics of tidal rivers by artificial neural networks and genetic algorithm

Authors: Arash Adib; Mohammad Nasiriyani

Addresses: Civil Engineering Department, Engineering Faculty, Shahid Chamran University, Ahvaz, Iran ' Civil Engineering Department, Engineering Faculty, Shahid Chamran University, Ahvaz, Iran

Abstract: Interaction of fluvial and tidal flows is a complex phenomenon in tidal rivers study. For determination of tidal characteristics in tidal rivers, an appropriate method is needed. Artificial neural network (ANN) using genetic algorithm (GA) method for determining the parameters is a suitable tool that can determine tide velocity, ebb velocity and variation of water surface elevation by tidal flow. Learning rates and momentum constants are optimised in ANN with GA method. It is observed that results of trained ANN by GA are more accurate than results of ANN. GA increases regression coefficient between output of ANN and desired output from 1% to 8% also decreases MSE extremely (43% to 85%). For case study, the Karun River was selected. The selected reach is between Ahvaz upstream and junction of three branches of Khoramshar downstream.

Keywords: artificial neural networks; ANNs; back propagation; fluvial flow; genetic algorithms; Karun River; tidal characteristics; Iran; tidal rivers; tidal flow; tide velocity; ebb velocity; water surface elevation.

DOI: 10.1504/IJW.2016.073739

International Journal of Water, 2016 Vol.10 No.1, pp.13 - 27

Received: 15 Oct 2013
Accepted: 13 May 2014

Published online: 17 Dec 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article