TY - JOUR ID - 85467 TI - Estimating the Fall Velocity of Sediment Particles Using Artificial Neural Network JO - Journal of Hydraulics JA - JHYD LA - en SN - 2345-4237 AU - Sadat-Helbar, S.M. AU - Amiri-Tokaldany, E. AU - Darzi, F. AD - Y1 - 2008 PY - 2008 VL - 3 IS - 2 SP - 59 EP - 65 DO - 10.30482/jhyd.2008.85467 N2 - The fall velocity of sediment particles is one of the important parameters in the phenomenon ofsediment transport, river bed and bank morphology, reservoir sedimentation and designing settlingbasins of water transport networks. To estimate the sediment fall velocity, many relationships in theliterature have been used by scientists and engineers but they have limitations. In this research, usingan Artificial Neural Network, a model to estimate the sediment fall velocity is introduced. The modelis designed and validated using 115 series of data presented in different researches covering anextensive range of sediment and fluid characteristics. The multi layer perception network with quickback propagation learning scheme is used to estimate the nonlinear mapping between input data, i.e.independent variables, and the output of the network, i.e. dependent variable. This nonlinear mappingis used to estimate the fall velocity. To evaluate prediction accuracy of the model, predictions of thedesigned network are compared with 14 experimental data set and analytical models of previousresearches. Comparisons were made using different error measures and it is found that the predictionaccuracy of the artificial network model is better than existing models. UR - https://jhyd.iha.ir/article_85467.html L1 - https://jhyd.iha.ir/article_85467_989045adcc01ed81d4c6a8e4aba6b98b.pdf ER -