基于VMD与优化LightGBM的混凝土拱坝变形预测  

Deformation Prediction of Concrete Arch Dam Based on VMD and Optimized LightGBM

作　　者：董志豪 赵二峰[1] 刘峰 宋桂华 吴斌庆 黎祎 DONG Zhi-hao;ZHAO Er-feng;LIU Feng;SONG Gui-hua;WU Bin-qing;LI Yi(College of Water Conservancy and Hydropower Engineering,Hohai University,Nanjing 210098,China;Shanghai Investigation,Design and Research Institute Co.,Ltd.,Shanghai 200434,China)

机构地区：[1]河海大学水利水电学院,江苏南京210098 [2]上海勘测设计研究院有限公司,上海200434

出　　处：《水电能源科学》2024年第8期132-136,共5页Water Resources and Power

基　　金：国家自然科学基金项目(52079046,U2243223)。

摘　　要：变形是反映混凝土拱坝安全运行状态的重要指标,因此变形预测模型的研究对拱坝结构健康评价具有重要意义。为充分挖掘拱坝变形监测数据的有效信息,提高监控模型的预测精度,提出一种基于变分模态分解与优化LightGBM的混凝土拱坝变形预测模型。首先,采用VMD将变形实测数据分解为多个模态分量;其次,引入改进灰狼算法与LightGBM相结合建立混凝土拱坝变形预测模型;随后,对模态分量进行单独建模和预测,最后叠加以得到最终的预测结果。工程实例分析表明,通过有效地分解重构,构建的变形预测模型具有较高的预测精度和泛化性能。Deformation is an important index to reflect the safe operation state of concrete arch dams.Therefore,the study of deformation prediction model is of great significance to the structural health evaluation of arch dams.In order to fully excavate the effective information of arch dam deformation monitoring data and improve the prediction accuracy of monitoring model,a deformation prediction model of concrete arch dam based on variational mode decomposition and optimization LightGBM is proposed.Firstly,the measured deformation data is decomposed into multiple modal components by using VMD.Secondly,the improved grey wolf algorithm combined with LightGBM is introduced to build the deformation prediction model of concrete arch dam.Then,the modal components are modeled and predicted respectively,and the final prediction result is obtained by superposition.The analysis of engineering examples shows that the deformation prediction model constructed by effective decomposition and reconstruction has high prediction accuracy and generalization performance.

关 键 词：变形预测 变分模态分解 改进灰狼算法 轻量梯度提升机 

分 类 号：TV698.11[水利工程—水利水电工程]