Efficiency assessment of local prediction method considering reconstruction of phase space and artificial neural network model for prediction of runoff (Case Study: Pole-kohneh Station, Kermanshah)

مهندسی عمران فردوسی

Article 7, Volume 28, Issue 2 - Serial Number 16, April 2016, Pages 91-108   

Document Type: پژوهشی

DOI: 10.22067/civil.v28i2.41135

Authors

Mohammad Zounemat-Kermani^{* 1}; Khatereh Amirkhani²

¹Shahid Bahonar university of Kerman

²Shahid Bahonar University of Kerman

Abstract

In this research prediction methods of artificial neural network and chaos theory are employed to predict daily, weekly and monthly runoff. For this, runoff series data observed at Pole-Kohneh located in the Qareh-Soo River. The nonlinear predictions of chaos are found to be in close agreement with the observed runoff, with high correlation coefficient for daily and weekly time scales. Predicted results of monthly time scale are not satisfying which indicating the chaos behavior in daily and weekly scales. The predicted results of ANN are inferior to chaos for daily and weekly scales but superior to chaos for monthly scale.

Keywords

Runoff time series; Chaos Theory; local prediction method

Supplementary Files

7.pdf