水利信息化Issue(3)：31-34,4.DOI:10.19364/j.1674-9405.2025.03.005
基于VMD-LSTM模型的王家坝月径流预测应用研究
Application research on monthly runoff prediction of Wangjiaba based on VMD-LSTM model
摘要
Abstract
In response to the issues of an incomplete water resource monitoring system,low accuracy of runoff prediction,and insufficient intelligence in water resource management and allocation to meet the requirements of the"Four Pre"business application in the Huaihe River Basin,a VMD-LSTM model was constructed based on VMD and the LSTM.VMD was used to decompose the original runoff sequence,while LSTM captured the long-term dependency relationships of subsequences.In the application of VMD-LSTM model to runoff prediction for the Wangjiaba section of the Huaihe River,compared with BP and LSTM models,the prediction results showed significant improvements in root mean square error,mean absolute error,and determination coefficient evaluation indicators.Based on monthly runoff prediction results of Wangjiaba Station on the Huaihe River,VMD-LSTM model effectively identified the non-stationary characteristics of runoff sequences,learns long-term dependencies,improved the accuracy of monthly runoff prediction.The model provides scientific and intelligent decision support for watershed water resource management and allocation.关键词
月径流预测/VMD/LSTM/组合模型Key words
monthly runoff prediction/VMD/LSTM/combination model分类
水利科学引用本文复制引用
鲁志杰,杨丰源,杜昀宸,胡友兵,赵梦杰,陈邦慧..基于VMD-LSTM模型的王家坝月径流预测应用研究[J].水利信息化,2025,(3):31-34,4.基金项目
国家重点研发计划项目(2024YFC3212800) （2024YFC3212800）
安徽省自然科学基金项目(2208085UQ10) （2208085UQ10）
