江淮水利科技2026，Vol.21Issue(1)：16-20,5.DOI:10.20011/j.cnki.JHWR.202601003
基于机器学习组合模型的引水工程高边坡位移预测研究
Displacement prediction of high slopes in water diversion projects based on ensemble machine learning models
王建军 1韩炎 1黄铭 2杨厚岗 2杨帆2
作者信息
- 1. 安徽省水利水电勘测设计研究总院股份有限公司,安徽 合肥 230088
- 2. 合肥工业大学土木与水利工程学院,安徽 合肥 230009
- 折叠
摘要
Abstract
To establish a reliable displacement prediction model for monitoring and early warning of high slopes in water diversion projects,the NeuralProphet model was introduced into high slope displacement prediction.Its trend module and autoregressive module were utilized to accurately learn the displacement patterns of high slopes in water diversion projects.The Optuna framework was employed to optimize the hyperparameters of the NeuralProphet model,leading to the development of the Optuna-NeuralProphet high slope displacement prediction model for water diversion projects,achieving accurate prediction of high slope displacement patterns.Taking the high slope in the Jianghuai communication section of the Yangtze-to-Huaihe water diversion project as the research subject,displacement monitoring data were trained and predicted,and comparative validation was conducted with the NeuralProphet model.The results indicated that the constructed Optuna-NeuralProphet model exhibited high accuracy and can more effectively predict displacements,providing a new approach and method for high slope displacement prediction in water diversion projects.关键词
位移预测/引水工程高边坡/NeuralProphet模型/Optuna框架Key words
displacement prediction/high slope in diversion project/NeuralProphet model/Optuna framework分类
建筑与水利引用本文复制引用
王建军,韩炎,黄铭,杨厚岗,杨帆..基于机器学习组合模型的引水工程高边坡位移预测研究[J].江淮水利科技,2026,21(1):16-20,5.
