Abstract
The short-term streamflow forecast is an important parameter in studies related to energy generation and the prediction of possible floods. Flowing through three Brazilian states, the Paraíba do Sul river is responsible for the supply and energy generation in several municipalities. Machine learning techniques have been studied with the aim of improving these predictions through the use of hydrological and hydrometeorological parameters. Furthermore, the predictive performance of the machine learning techniques are directly related to the quality of the training base and, moreover, to the set of hyperparameters used. The present study explores the combination of the Gradient Boosting technique coupled with a Genetic Algorithm to found the best set of hyperparameter to maximize the predicting performance of the Paraíba do Sul river streamflow.
Highlights
Natural discharges are important for various purposes of use in activities that accompany the development of humanity and predicting river water flow is vital to water management tasks, such as improving the efficiency of hydroelectric power generation, agricultural management, and flood control
This paper proposes to use computational intelligence techniques combined with evolutionary algorithms to forecast short-term flows to support decision-makers regarding the use of water resources in the Paraíba do Sul river basin
The forecast of the flow of the Paraíba do Sul river is performed for 7 subsequent days, based on the precipitation and flow information collected in the previous 14 days
Summary
Natural discharges are important for various purposes of use in activities that accompany the development of humanity and predicting river water flow is vital to water management tasks, such as improving the efficiency of hydroelectric power generation, agricultural management, and flood control. Hydroelectric power generation in the short term (up to seven days) using modeling techniques and computational learning to predict flows in watercourses. Models for streamflow prediction depend on parameters whose values need to be identified in order to provide good estimations Optimization techniques, such as nature-inspired algorithms, are an alternative to search towards good parameter sets. In spite of the application of several machine learning techniques available in the literature, the Gradient Boosting (GB) approach (FRIEDMAN, 2000) has not been widely applied to predicting daily streamflows and has been little discussion on parameter selection for GB. This paper proposes to use computational intelligence techniques combined with evolutionary algorithms to forecast short-term flows to support decision-makers regarding the use of water resources in the Paraíba do Sul river basin.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Revista Mundi Engenharia, Tecnologia e Gestão (ISSN: 2525-4782)
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
