Next Article in Journal
Experimental Investigation of Chlorella vulgaris and Enterobacter sp. MN17 for Decolorization and Removal of Heavy Metals from Textile Wastewater
Previous Article in Journal
Concentration–Discharge Relationships in Runoff Components during Rainfall Events at the Hydrohill Experimental Catchment in Chuzhou, China
Article

Prediction of Streamflow Based on Dynamic Sliding Window LSTM

1
Upper Changjiang River Bureau of Hydrological and Water Resources Survey, Changjiang Water Resources Commission, Chongqing 400020, China
2
School of Computer Science and Engineering, Xi’an University of Technology, Xi’an 710048, China
3
Department of Applied Informatics, Vytautas Magnus University, 44404 Kaunas, Lithuania
4
Department of Intelligent Computer Systems, Częstochowa University of Technology, 42200 Częstochowa, Poland
5
Faculty of Applied Mathematics, Silesian University of Technology, 44100 Gliwice, Poland
*
Author to whom correspondence should be addressed.
Water 2020, 12(11), 3032; https://doi.org/10.3390/w12113032
Received: 7 September 2020 / Revised: 13 October 2020 / Accepted: 13 October 2020 / Published: 29 October 2020
The streamflow of the upper reaches of the Yangtze River exhibits different timing and periodicity characteristics in different quarters and months of the year, which makes it difficult to predict. Existing sliding window-based methods usually use a fixed-size window, for which the window size selection is random, resulting in large errors. This paper proposes a dynamic sliding window method that reflects the different timing and periodicity characteristics of the streamflow in different months of the year. Multiple datasets of different months are generated using a dynamic window at first, then the long-short term memory (LSTM) is used to select the optimal window, and finally, the dataset of the optimal window size is used for verification. The proposed method was tested using the hydrological data of Zhutuo Hydrological Station (China). A comparison between the flow prediction data and the measured data shows that the prediction method based on a dynamic sliding window LSTM is more accurate by 8.63% and 3.85% than the prediction method based on fixed window LSTM and the dynamic sliding window back-propagation neural network, respectively. This method can be generally used for the time series data prediction with different periodic characteristics. View Full-Text
Keywords: streamflow; flow prediction; dynamic sliding window; deep learning; neural network; LSTM streamflow; flow prediction; dynamic sliding window; deep learning; neural network; LSTM
Show Figures

Figure 1

MDPI and ACS Style

Dong, L.; Fang, D.; Wang, X.; Wei, W.; Damaševičius, R.; Scherer, R.; Woźniak, M. Prediction of Streamflow Based on Dynamic Sliding Window LSTM. Water 2020, 12, 3032. https://doi.org/10.3390/w12113032

AMA Style

Dong L, Fang D, Wang X, Wei W, Damaševičius R, Scherer R, Woźniak M. Prediction of Streamflow Based on Dynamic Sliding Window LSTM. Water. 2020; 12(11):3032. https://doi.org/10.3390/w12113032

Chicago/Turabian Style

Dong, Limei; Fang, Desheng; Wang, Xi; Wei, Wei; Damaševičius, Robertas; Scherer, Rafał; Woźniak, Marcin. 2020. "Prediction of Streamflow Based on Dynamic Sliding Window LSTM" Water 12, no. 11: 3032. https://doi.org/10.3390/w12113032

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Search more from Scilit
 
Search
Back to TopTop