基于随机森林与CNN-GRU模型的大坝变形监测数据缺失值填补方法OA

In the process of dam operation,the missing data of deformation monitoring seriously affects the prediction and judgment of dam safety status.Currently,it is difficult to effectively consider the correlation between measurement points in the commonly used missing value filling methods,and it is difficult to fill the effect to meet the needs of safety monitoring.This paper proposes a method to fill the missing values of dam deformation monitoring data based on Random Forest and CNN-GRU model.Firstly,a random forest algorithm is used to analyze the correlation between measured values and loads and between measured values and measured values,screen out the environmental factors that have a greater impact on measured values,and group the measurement points with strong correlation into one class to construct a multi-measurement point safety monitoring model.Then,on this basis,a convolutional neural network(CNN)and a gated recurrent unit neural network(GRU)model are trained,and a CNN-GRU-based missing value filling method for dam deformation monitoring is proposed to realize the automated filling of missing values at multiple measurement points.The validity of the proposed method is verified by case analysis,which lays a foundation for scientific assessment of the serviceability of dams.