Year: 2026
Type: article
Abstract: This study employed Gaussian Process Regression (GPR) and Adaptive Neuro-Fuzzy Inference Systems (ANFIS) to predict sediment transport in circular pipes for water and wastewater systems. It evaluated the influence of hydraulic parameters and sediment particle characteristics on sediment discharge. Both methods provided reliable and accurate predictions, with models incorporating hydraulic radius, relative particle size, and friction coefficient performing best. The model using inputs (λ s , F d , D gr , d50/R) achieved the highest accuracy in both rough and smooth pipe conditions. Sensitivity analysis highlighted the particle Froude number (F d ) as the most critical factor affecting sediment discharge. Smoother pipe beds yielded more accurate predictions than rougher ones, as increased roughness reduced sediment transport efficiency. The lowest bed roughness (0.53 mm) resulted in the most precise predictions.
Author: Ramin Vafaei Poursorkhabi
Institution: Islamic Azad University of Tabriz
Language: English
Subfield: Ecology
Field: Environmental Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,490
OpenAlex ID: W7128743442