Research on dam deformation prediction based on variable weight combination model

In order to establish a dam deformation prediction model and method that is more consistent with dam deformation and achieve higher accuracy, a dam deformation prediction method based on variable weight combination model is proposed. First, four popular machine learning prediction models are selected to predict dam deformation, and LightGBM model and LSTM are obtained Then, combined with the characteristics of each model, genetic algorithm is used to assign weight coefficients to the combined model at different times, and then the predicted value of the final deformation of the dam is obtained by combining the prediction results of the two dominant models. Finally, according to the measured deformation data of a dam, an application example is analyzed. The results show that the combined forecasting model can combine the advantages of different forecasting models, and the prediction accuracy and reliability are obviously higher than that of the single forecasting model, and it has strong practicability.