Dam deformation monitoring data prediction using Bayesian optimization Gaussian process regression

In order to ensure the safe operation of dams, it is necessary to carry out safety monitoring, among which deformation monitoring is the most important.A large number of deformation monitoring data should be analyzed to determine the safety state of the dam. Aiming at constructing a high accuracy dam deformation prediction model, Gaussian process regression(GPR) model which is a widely used data prediction model has the high adaptability to non-stationary data.The selection of hyper-parameter is the key to the accuracy of GPR.In order to improve the prediction accuracy of GPR, Bayesian optimization(BO) is used to optimize its hyper-parameter.Compared the precition results of BP ,GPR and BO-GPR, the result shows that the Bayesian optimization Gaussian process regression model has higher prediction accuracy and is an effective method for dam deformation analysis and prediction.