Abstract: In mining areas,the surface movement and deformation are related to the safety of production
and the safety of ground buildings(structures). It is of great practical significance to regularly monitor and predict the surface movement and deformation in mining areas. Due to the long duration of surface subsidence in the mining areas,the observation data has the characteristics of long time series and nonlinear change,so there are some problems such as difficulty in the selection of modeling data interval and low prediction accuracy of GM(1,1). In view of the above problems,this paper establishes an improved GM(1,1)model based on the residual weighted correction,and puts forward the method of interval selection of modeling data, prediction process and evaluation standard of model accuracy. It is verified by an example that the prediction accuracy of GM(1,1)model is better than that of GM(1,1) model,and the method of interval selection of modeling data from stage to tail can improve the prediction accuracy.