Studies on dam deformation prediction and warning based on an intelligent combination model
Li Shuangping
Liu Zuqiang
Zhang Bin
Zheng Junxing
Wang Huawei
Li Yonghua
Su Sennan
Abstract：To satisfy the requirements of accuracy and reliability of"forecasting,early-warning,rehearsal and emergency planning"for dam safety in digital twin water resources project construction,an intelligent combination model was developed.This method separates the dominant trend component of dam deformation by evaluating multiple influential factors and combining signal processing technology.Intelligent algorithms were then used to accurately match the optimal fitting model.Various modeling technologies including grey model,time series model and neural networks were adopted to build a highly integrated and adaptive intelligent combination model.Through training and optimization of deformation time series of the Danjiangkou Dam,and comparing with the prediction results of traditional statistical models,the experiment indicates that the intelligent combination model has significant advantages in prediction accuracy,data adaptability and robustness,especially in dealing with nonlinear relationships and long-term dependencies.At the same time,it effectively improves the accuracy of extended prediction and generalization ability.In addition,the model can accurately predict the potential deformation trend of key parts of the dam one cycle in advance(with a duration of one year),providing sufficient time for prevention and risk reduction.Application of the model for deformation prediction and forecasting of the Danjiangkou Dam provides a strong support for dam safety assessment,risk warning and scientific management.
Keywords：intelligent combination modelgeneralization abilityrobustnessdam deformationprediction and forecasting
Publication Date：2025-01-26
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：8( 65-72 )