Author(s): Junchao Shi; Xinjie Li; Hailong Wang; Xiaofei Yan; Qiang Wang; Jie Liu

Linked Author(s): Jie Liu

Keywords: Machine learning algorithm; Reservoir sediment discharge; Prediction model

Abstract: In response to the limited capability of one-dimensional flow and sediment models to represent complex hydrological conditions, this research incorporates two machine learning algorithms, namely K-Nearest Neighbors (KNN) and Gradient Boosting Regression (GBR), to predict sediment discharge in reservoirs. The simulation utilizes flow and sediment data from the Xiaolangdi Reservoir spanning the years 2000 to 2019. The simulation results indicate that among the various machine learning algorithms employed for constructing sediment discharge prediction models, the Gradient Boosting Regression (GBR) exhibits the best performance. The experimental findings demonstrate that the algorithmic simulation results can serve as references for sediment discharge prediction in reservoirs and subsequent scheduling strategies.

DOI: https://doi.org/10.3850/iahr-hic2483430201-280

Year: 2024