Predicting longitudinal dispersion coefficient in natural streams by artificial neural network

被引:90
作者
Tayfur, G [1 ]
Singh, VP
机构
[1] Izmir Inst Technol, Fac Engn, Dept Civil Engn, TR-35340 Izmir, Turkey
[2] Louisiana State Univ, Dept Civil & Environm Engn, Baton Rouge, LA 70803 USA
关键词
D O I
10.1061/(ASCE)0733-9429(2005)131:11(991)
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
An artificial neural network (ANN) model was developed to predict the longitudinal dispersion coefficient in natural streams and rivers. The hydraulic variables [flow discharge (Q), flow depth (H), flow velocity (U), shear velocity (u*), and relative shear velocity (U/u*)] and geometric characteristics [channel width (B), channel sinuosity (sigma), and channel shape parameter (beta)] constituted inputs to the ANN model, whereas the dispersion coefficient (K-x) was the target model output. The model was trained and tested using 71 data sets of hydraulic and geometric parameters and dispersion coefficients measured on 29 streams and rivers in the United States. The training of the ANN model was accomplished with an explained variance of 90% of the dispersion coefficient. The dispersion coefficient values predicted by the ANN model satisfactorily compared with the measured values corresponding to different hydraulic and geometric characteristics. The predicted values were also compared with those predicted using several equations that have been suggested in the literature and it was found that the ANN model was superior in predicting the dispersion coefficient. The results of sensitivity analysis indicated that the Q data alone would be sufficient for predicting more frequently occurring low values of the dispersion coefficient (K-x < 100 m(2)/s). For narrower channels (B/H < 50) using only U/u* data would be sufficient to predict the coefficient. If beta and sigma were used along with the flow variables, the prediction capability of the ANN model would be significantly improved.
引用
收藏
页码:991 / 1000
页数:10
相关论文
共 37 条
[1]  
[Anonymous], 1973, J HYDRAUL DIV
[2]   A NEURAL-NETWORK APPROACH TO THE DETERMINATION OF AQUIFER PARAMETERS [J].
AZIZ, ARA ;
WONG, KFV .
GROUND WATER, 1992, 30 (02) :164-166
[3]  
Chang H.H., 1988, Fluvial Processes in River Engineering
[4]   Hybrid neural network-finite element river flow model [J].
Chua, LHC ;
Holz, KP .
JOURNAL OF HYDRAULIC ENGINEERING, 2005, 131 (01) :52-59
[5]   An artificial neural network approach to rainfall-runoff modelling [J].
Dawson, CW ;
Wilby, R .
HYDROLOGICAL SCIENCES JOURNAL-JOURNAL DES SCIENCES HYDROLOGIQUES, 1998, 43 (01) :47-66
[6]   Longitudinal dispersion coefficient in single-channel streams [J].
Deng, ZQ ;
Bengtsson, L ;
Singh, VP ;
Adrian, DD .
JOURNAL OF HYDRAULIC ENGINEERING-ASCE, 2002, 128 (10) :901-916
[7]   Longitudinal dispersion coefficient in straight rivers [J].
Deng, ZQ ;
Singh, VP ;
Bengtsson, L .
JOURNAL OF HYDRAULIC ENGINEERING-ASCE, 2001, 127 (11) :919-927
[8]   Artificial neural networks for streamflow prediction [J].
Dolling, OR ;
Varas, EA .
JOURNAL OF HYDRAULIC RESEARCH, 2002, 40 (05) :547-554
[9]   THE DISPERSION OF MARKED FLUID IN TURBULENT SHEAR FLOW [J].
ELDER, JW .
JOURNAL OF FLUID MECHANICS, 1959, 5 (04) :544-560
[10]  
Fischer H. B., 1967, J HYDR DIV ASCE, V93, P187, DOI DOI 10.1061/JYCEAJ.0001706