Short-term streamflow forecasting using the feature-enhanced regression model
ID Bai, Yun (Author), ID Bezak, Nejc (Author), ID Lebar, Klaudija (Author), ID Klun, Mateja (Author), ID Zhang, Jin (Author)

.pdfPDF - Presentation file, Download (1,22 MB)
MD5: A7D13E36EDE6F1D9749C40844064313D
URLURL - Source URL, Visit https://link.springer.com/content/pdf/10.1007%2Fs11269-019-02399-1.pdf This link opens in a new window

Reservoir inflow forecasting is extremely important for the management of a reservoir. In practice, accurate forecasting depends on the feature learning performance. To better address this issue, this paper proposed a feature-enhanced regression model (FER), which combined stack autoencoder (SAE) with long short-term memory (LSTM). This model had two constituents: (1) The SAE was constructed to learn a representation as close as possible to the original inputs. Through deep learning, the enhanced feature could be captured sufficiently. (2) The LSTM was established to simulate the mapping between the enhanced features and the outputs. Under recursive modeling, the patterns of correlation in the short term and dependence in the long term were considered comprehensively. To estimate the performance of the FER model, two historical daily discharge series were investigated, i.e., the Yangtze River in China and the Sava Dolinka River in Slovenia. The proposed model was compared with other machine-learning methods (i.e., the LSTM, SAE-based neural network, and traditional neural network). The results demonstrated that the proposed FER model yields the best forecasting performance in terms of six evaluation criteria. The proposed model integrates the deep learning and recursive modeling, and thus being beneficial to exploring complex features in the reservoir inflow forecasting. Moreover, for smaller catchments with significant torrential characteristics, more data are needed (e.g., at least 20 years) to effectively train the model and to obtain accurate flood-forecasting results.

Keywords:long short-term memory, stack autoencoder, feature enhanced, daily reservoir inflow, forecast
Work type:Scientific work
Typology:1.01 - Original Scientific Article
Organization:FGG - Faculty of Civil and Geodetic Engineering
Publication status in journal:In print
Article version:Postprint, final article version, accepted into publication
Publisher:Springer Nature B. V.
Number of pages:Str. 4783-4797
Numbering:Letn.33, št. nov.
PID:20.500.12556/RUL-114378 This link opens in a new window
ISSN on article:0920-4741
DOI:10.1007/s11269-019-02399-1 This link opens in a new window
COBISS.SI-ID:8969057 This link opens in a new window
Publication date in RUL:24.02.2020
Copy citation
AddThis uses cookies that require your consent. Edit consent...

Record is a part of a journal

Title:Water resources management
Shortened title:Water resour. manag.
COBISS.SI-ID:512526105 This link opens in a new window


License:CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
Description:The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.
Licensing start date:24.02.2020

Secondary language

Keywords:hidrotehnika, dolgotrajni kratkoročni spomin, avtoenkoder, vhodna spremenljivka, dnevni vtok v rezervoar, napovedi


Funder:ARRS - Agencija za raziskovalno dejavnost Republike Slovenije
Project number:P2-0180
Name:Vodarstvo in geotehnika: orodja in metode za analize in simulacije procesov ter razvoj tehnologij.

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections: