Flood water level modeling and prediction using NARX neural network: Case study at Kelang river

Abstract

Flood disaster has becomes major threat around the world because it causes loss of lives and damages to property. Thus, reliable flood prediction is very much needed in order to reduce the effects of flood disaster. Hence, an accurate flood water level prediction is an important task to achieve. Since flood water level fluctuation is highly nonlinear, it is very difficult to predict the flood water level. Artificial Neural Network is well known technique is solving nonlinear cases and Nonlinear Auto Regressive with Exogenous Input (NARX) model is one class of Artificial Neural Network model. Thus, this paper proposes flood water level modeling and prediction using Nonlinear Auto Regressive with Exogenous Input (NARX) model to overcome the nonlinearity problem and come out with an advanced neural network model for the prediction of flood water level 10 hours in advance. The input and output parameters used in this model are based on real-time data obtained from Department of Irrigation and Drainage Malaysia. Results showed that NARX model successfully predicted the flood water level 10 hours ahead of time.