7 hours flood prediction modelling using NNARX structure: Case study Terengganu

Most of the countries have paid great attention to flood water level prediction since flood may damages people's life and property. Currently, hydrological models were used to get the prediction of flood water levels. However, this involved with various parameters such as hydrometric measurements, weather forecasts and hydrogeological maps, in addition to water level, temperature and flow observations. Therefore, such models are usually difficult to develop especially when describing large and complex system such as the dynamic of flood water level. Since flood water level fluctuate highly nonlinear, it is very difficult to predict the flood water level. Since Artificial Neural Network was proven to be best model to handle nonlinear cases, this paper proposed flood prediction modelling using Artificial Neural Network (ANN) technique with 7 hours prediction time. The area of study was Terengganu where the input parameters used in this modelling were river water level at upstream stations whereas output parameter was river water level at downstream station or so called flood location. 542 samples data collected from 15/12/2011 till 19/12/2011 were used for modelling, 542 samples data collected from 26/2/2012 till 1/3/2011 were used for model validation and 428 samples data collected data from 4/6/2013 till 7/6/2013 were used for model testing. Results showed that NNARX model successfully predicted flood water level 7 hours ahead of time.

[1]  Ahmad Golbabai,et al.  TIME SERIES PREDICTION USING WIDTH SCALING IN RBF NETWORKS , 2013 .

[2]  Tharam S. Dillon,et al.  Neural-Network-Based Models for Short-Term Traffic Flow Forecasting Using a Hybrid Exponential Smoothing and Levenberg–Marquardt Algorithm , 2012, IEEE Transactions on Intelligent Transportation Systems.

[3]  Jun Wang,et al.  Financial time series prediction by a random data-time effective RBF neural network , 2014, Soft Comput..

[4]  Mohammad Taghi Hamidi Beheshti,et al.  A local linear radial basis function neural network for financial time-series forecasting , 2010, Applied Intelligence.

[5]  Noureddine Zerhouni,et al.  Defining and applying prediction performance metrics on a recurrent NARX time series model , 2010, Neurocomputing.

[6]  Carl E. Rasmussen,et al.  Automated Bayesian System Identification with NARX Models , 2013, ArXiv.

[7]  U. C. Kothyari,et al.  Modeling of the daily rainfall-runoff relationship with artificial neural network , 2004 .

[8]  Guilherme De A. Barreto,et al.  Long-term time series prediction with the NARX network: An empirical evaluation , 2008, Neurocomputing.

[9]  Jonghwa Ham,et al.  Dynamic Baysesian state-space model with a neural network for an online river flow prediction , 2013 .

[10]  Shalini Bhatia,et al.  Traffic Flow Control using Neural Network , 2012 .

[11]  A. Shamseldin.  Application of a neural network technique to rainfall-runoff modelling , 1997 .

[12]  Saeed Zolfaghari,et al.  Chaotic time series prediction with residual analysis method using hybrid Elman-NARX neural networks , 2010, Neurocomputing.

[13]  L. Re,et al.  Iterative Identification of Polynomial NARX Models for Complex Multi-Input Systems , 2010 .

[14]  Li-Chiu Chang,et al.  Online multistep-ahead inundation depth forecasts by recurrent NARX networks , 2012 .

[15]  Karim C. Abbaspour,et al.  A wavelet-neural network hybrid modelling approach for estimating and predicting river monthly flows , 2013 .

[16]  K. Sudheer,et al.  Constructing prediction interval for artificial neural network rainfall runoff models based on ensemble simulations , 2013 .

[17]  A. K. Lohani,et al.  Comparative study of neural network, fuzzy logic and linear transfer function techniques in daily rainfall‐runoff modelling under different input domains , 2011 .

[18]  Marco Lovera,et al.  Identification of the rainfall-runoff relationship in urban drainage networks , 1999 .