Suspended sediment load estimation using Support Vector Machine and improved computational methods

Articles in Press

Document Type : Research Paper

Authors

1 Bu-Ali Sina university

2 Department of Water Sciences and Engineering, Faculty of Agriculture, Bu-Ali Sina University, Hamedan, Iran

3 Department of Statistics, Faculty of Science, Payame Noor University, Tehran, Iran

Abstract

Accurately estimating the suspended sediment load of a river is crucial for the study and development of water resource systems. However, direct measurement of sediment load is expensive and time-consuming. Therefore, developing alternative models, such as gauge curves or simulation models, is recommended. This study employed two methods to predict the river's suspended sediment load. The first method used a combination of a support vector machine (SVM) model and a refrigeration algorithm (SA), while the second method used a gauge curve and a modified FAO sediment measurement method. To validate the modeling results, water flow and suspended sediment load data from Yelfan River between 2006-2014 were used. Normalizing the input data to the SVM model improved the standard error index (SE) by 9.93% in the modeling test section. Additionally, modifying the input pattern from single input-single output (SISO) to multiple input-single output (MISO) improved this index by 1.03%. After optimizing the calibration curve coefficients and the FAO correction method using the general reduced gradient (GRG), the SE index improved by 10.13% and 73.19%, respectively. Among the reviewed methods, the SVM model based on the normalized multi-input-single-output model was found to be the best method for predicting the suspended sediment load of the river.

Keywords

Main Subjects

Irrigation Sciences and Engineering

Articles in Press, Accepted Manuscript
Available Online from 21 October 2024

Files

History

  • Receive Date: 29 September 2023
  • Revise Date: 20 August 2024
  • Accept Date: 21 October 2024
  • Publish Date: 21 October 2024

Share

How to cite

Statistics