Transformer-Based Methods for Water Level Prediction: A Case Study of the Kien Giang River, Quang Binh Province
Bao Bui Quoc1, Hung Nguyen Khanh2, Hieu Nguyen Dac3, Dat Tran-Anh4, Quang Chieu Ta5
1Bao Bui Quoc, Hanoi University of Science and Technology, No. 1 Dai Co Viet, Hai Ba Trung, 100000, Hanoi, Vietnam.
2Hung Nguyen Khanh, Hanoi University of Science and Technology, No. 1 Dai Co Viet, Hai Ba Trung, 100000, Hanoi, Vietnam.
3Hieu Nguyen Dac, Department of Artificial Intelligence, Thuyloi University, 175 Tay Son, Dong Da, 100000, Hanoi, Vietnam.
4Dat Tran Anh, Department of Artificial Intelligence, Thuyloi University, 175 Tay Son, Dong Da, 100000, Hanoi, Vietnam.
5Quang Chieu Ta, Department of Artificial Intelligence, Thuyloi University, 175 Tay Son, Dong Da, 100000, Hanoi, Vietnam.
Manuscript received on 30 June 2024 | Revised Manuscript received on 08 July 2024 | Manuscript Accepted on 15 July 2024 | Manuscript published on 30 July 2024 | PP: 21-28 | Volume-13 Issue-8, July 2024 | Retrieval Number: 100.1/ijitee.H993613080724 | DOI: 10.35940/ijitee.H9936.13080724
Open Access | Editorial and Publishing Policies | Cite | Zenodo | OJS | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Accurate and timely water level prediction is of paramount importance in various applications, including flood forecasting, hydroelectric power management, and environmental monitoring. Traditional recurrent neural network (RNN)-based methods have been widely used for this task. However, recent advancements in long-term time-series forecasting have introduced transformer-based models that have significantly improved the performance in time-series prediction tasks. In this research, we investigate the application of transformer-based models to the task of water level prediction, specifically focusing on the Nhat Le River Basin. We conducted multiple experiments with different test cases and various model architectures, providing specific analyses of the model’s prediction capabilities. The transformer-based models consistently outperformed conventional RNN-based methods across a range of evaluation metrics, including root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). Moreover, these models exhibited excellent flood peak prediction accuracy, with errors consistently below 0.02 meters. The robustness and scalability of transformer-based models make them promising for accurate water-level predictions in real-world applications.
Keywords: Water level Prediction, Time Series Forecasting, Transformer-Based Models, Deep Learning.
Scope of the Article: Electrical and Electronics