Research Article
Flood Prediction Using Multilayer Perceptron Networks and Long Short-Term Memory Networks at Thu Bon-Vu Gia Catchment, Vietnam
@INPROCEEDINGS{10.1007/978-3-030-77424-0_32, author={Duy Vu Luu and Thi Ngoc Canh Doan and Khanh Le Nguyen and Ngoc Duong Vo}, title={Flood Prediction Using Multilayer Perceptron Networks and Long Short-Term Memory Networks at Thu Bon-Vu Gia Catchment, Vietnam}, proceedings={Industrial Networks and Intelligent Systems. 7th EAI International Conference, INISCOM 2021, Hanoi, Vietnam, April 22-23, 2021, Proceedings}, proceedings_a={INISCOM}, year={2021}, month={5}, keywords={Multilayer Perceptron (MLP) Long Short-Term Memory (LSTM) Thu Bon-Vu Gia catchment}, doi={10.1007/978-3-030-77424-0_32} }- Duy Vu Luu
Thi Ngoc Canh Doan
Khanh Le Nguyen
Ngoc Duong Vo
Year: 2021
Flood Prediction Using Multilayer Perceptron Networks and Long Short-Term Memory Networks at Thu Bon-Vu Gia Catchment, Vietnam
INISCOM
Springer
DOI: 10.1007/978-3-030-77424-0_32
Abstract
There is a significant change in the amplitude of rainfall between the rainy season and the dry season at Thu Bon-Vu Gia catchment in Vietnam. 65% to 80% of the annual rainfall is in the rainy season. Therefore, Thu Bon - Vu Gia catchment is a highly flood prone region. Floods frequently occur in this area and destroy critical infrastructure. This study compares Multilayer Perceptron (MLP) networks and Long Short-Term Memory (LSTM) networks in forecasting floods at Thu Bon-Vu Gia catchment. Discharges at the downstream point are predicted by utilizing periodic rainfall and flow data at upstream locations. Both models do not use other hydrologic, geological and meteorological data, which have low quality at the study site. Both models are reliable to forecast the flood in the catchment when the values of RMSE and NSE of the models are about 320 m3/s and 0.5 respectively.
