The Community for Technology Leaders
2016 International Conference on Frontiers of Information Technology (FIT) (2016)
Islamabad, Pakistan
Dec. 19, 2016 to Dec. 21, 2016
ISBN: 978-1-5090-5300-1
pp: 18-22
Fazlina Ahmat Ruslan , Faculty of Electrical Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, Malaysia
Abd Manan Samad , Centre of Surveying Science and Geomatics, Faculty of Arc., Planning and Surveying, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, Malaysia
Ramli Adnan , Faculty of Electrical Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, Malaysia
ABSTRACT
Flood prediction modelling is one of the most popular research areas among researcher around the world. This is due to negative impacts to the economy and society that were caused by flood. The dynamic behaviours of river water level that causing flood were commonly modelled by researcher either by equations using physic theories or by black-box model. River water level prediction model that could predict the occurrence of flood ahead of time was not developed or explored intensively. This paper proposed river water level prediction modelling for flood monitoring at Kuala Lumpur flood prone area using Neural Network Autoregressive with Exogenous Inputs (NNARX). Data used for modelling, validation and testing were previous event flood real-time water level data from the respective authority department. The modelling was carried out using Neural Network Toolbox of MATLAB. The prediction performance of NNARX model using 3 hours, 4 hours and 5 hours prediction time was presented. Results show that NNARX model with 4 hours prediction time achieved reliable prediction result.
INDEX TERMS
Black-Box Model, Flood Prediction Model, ANN, NNARX
CITATION
Fazlina Ahmat Ruslan, Abd Manan Samad, Ramli Adnan, "River water level prediction modelling for flood monitoring using NNARX", 2016 International Conference on Frontiers of Information Technology (FIT), vol. 00, no. , pp. 18-22, 2016, doi:10.1109/FIT.2016.7857511
86 ms
(Ver 3.3 (11022016))