No abstract

Abstract:In this study, three artificial neural network methods, i.e. feed forward back propagation, the radial basis function neural network, and the generalized regression neural network are employed to compute the longitudinal dispersion coefficient in order to evaluate its behaviour in predicting dispersion characteristics in natural streams. These methods, which use hydraulic and geometrical data to predict dispersion coefficients, can easily be applied to natural streams and are proven to be superior in explaining their dispersion characteristics more precisely than existing equations. This method of predicting the longitudinal dispersion coefficient in river flows was tested on 65 data sets, obtained by researchers from 30 rivers in the USA. Results using the models are compared with results obtained in many other studies, and are shown to be more accurate than the other methods considered.

No abstract

No abstract

Researchers have long used differential equations to investigate longitudinal dispersion processes, which can be derived under certain assumptions and include a longitudinal dispersion coefficient (D 1 ). In practice, most empirical equations are developed only for D 1 . Unfortunately, many critical assumptions in the derivation of these equations are not considered, and consequently, these equations can only be used with precautions and reservations. The goal of this study is to develop a fuzzy model to predict D 1 in natural channels. The model depends on 65 data sets extracted from the literature. The variables are the depth, width and mean cross-sectional velocity of the flow, shear velocity and D 1 . The data is divided for training and testing phases. The model is compared with measured data and seven existing equations. The comparison depends on seven statistical characteristics, four different error modes, and a contour map. It is observed that the fuzzy model yields results that are more reliable than existing methods and it can be used more easily and efficiently.