ANN Modeling for Prediction of Velocity in Channel Bends

Abstract

The flow in a channel bend is spiral or helical. It is a movement of water particles in the flow direction. Many researchers have stated mathematical equations to predict velocity in the flow direction in the channel bend. These equations are based on simplified assumptions. The ANN is a viable alternative to predict longitudinal velocity in channel bend. It builds the model by estimating suitable approximating function of the available input/output samples. Once such relationship is established & validated, it can be used for the prediction of the future system behavior. The paper aims at developing Artificial Neural Network model, namely Multilayer Perceptron (MLP). It is found that MLP model is capable to predict velocity in the channel bend with accuracy.