Abstract:In view of the "S" process of dam deformation,and the characteristics of small amount of deformation monitoring data, poor information and noise, the grey Verhulst model is used to fit and predict dam deformation. In view of that the noise reduces the smoothness of the data , and then affects the accuracy of the model,the original monitoring data is firstly smoothed by the function cot(x2) transform during modeling,and then the transformed data is used for modeling,Finally,the predicted value of the model is inversely transformed to obtain the final predicted value of dam deformation.Empirical analysis shows that cot (x2) - Grey Verhulst model has higher accuracy and better prediction effect than some existing models.