Next Article in Journal
Effects of Polymer Molecular Weight on Adsorption and Flocculation in Aqueous Kaolinite Suspensions Dosed with Nonionic Polyacrylamides
Next Article in Special Issue
Estimation of Rainfall Associated with Typhoons over the Ocean Using TRMM/TMI and Numerical Models
Previous Article in Journal
Water Discharge and Sediment Load Changes in China: Change Patterns, Causes, and Implications
Previous Article in Special Issue
Applying a Correlation Analysis Method to Long-Term Forecasting of Power Production at Small Hydropower Plants
Article Menu

Export Article

Open AccessArticle
Water 2015, 7(11), 5876-5895; doi:10.3390/w7115876

An Hourly Streamflow Forecasting Model Coupled with an Enforced Learning Strategy

1
Taiwan Typhoon and Flood Research Institute, National Applied Research Laboratories, 11F., No. 97, Sec. 1, Roosevelt Rd., Taipei City 10093, Taiwan
2
Department of Civil Engineering, National Taiwan University, Taipei City 10617, Taiwan
*
Author to whom correspondence should be addressed.
Academic Editor: Kwok-wing Chau
Received: 17 September 2015 / Revised: 19 October 2015 / Accepted: 22 October 2015 / Published: 28 October 2015
(This article belongs to the Special Issue Use of Meta-Heuristic Techniques in Rainfall-Runoff Modelling)
View Full-Text   |   Download PDF [933 KB, uploaded 28 October 2015]   |  

Abstract

Floods, one of the most significant natural hazards, often result in loss of life and property. Accurate hourly streamflow forecasting is always a key issue in hydrology for flood hazard mitigation. To improve the performance of hourly streamflow forecasting, a methodology concerning the development of neural network (NN) based models with an enforced learning strategy is proposed in this paper. Firstly, four different NNs, namely back propagation network (BPN), radial basis function network (RBFN), self-organizing map (SOM), and support vector machine (SVM), are used to construct streamflow forecasting models. Through the cross-validation test, NN-based models with superior performance in streamflow forecasting are detected. Then, an enforced learning strategy is developed to further improve the performance of the superior NN-based models, i.e., SOM and SVM in this study. Finally, the proposed flow forecasting model is obtained. Actual applications are conducted to demonstrate the potential of the proposed model. Moreover, comparison between the NN-based models with and without the enforced learning strategy is performed to evaluate the effect of the enforced learning strategy on model performance. The results indicate that the NN-based models with the enforced learning strategy indeed improve the accuracy of hourly streamflow forecasting. Hence, the presented methodology is expected to be helpful for developing improved NN-based streamflow forecasting models. View Full-Text
Keywords: streamflow forecasting; neural networks; support vector machine; enforced learning strategy streamflow forecasting; neural networks; support vector machine; enforced learning strategy
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. (CC BY 4.0).

Scifeed alert for new publications

Never miss any articles matching your research from any publisher
  • Get alerts for new papers matching your research
  • Find out the new papers from selected authors
  • Updated daily for 49'000+ journals and 6000+ publishers
  • Define your Scifeed now

SciFeed Share & Cite This Article

MDPI and ACS Style

Wu, M.-C.; Lin, G.-F. An Hourly Streamflow Forecasting Model Coupled with an Enforced Learning Strategy. Water 2015, 7, 5876-5895.

Show more citation formats Show less citations formats

Related Articles

Article Metrics

Article Access Statistics

1

Comments

[Return to top]
Water EISSN 2073-4441 Published by MDPI AG, Basel, Switzerland RSS E-Mail Table of Contents Alert
Back to Top