Abstract: In Changcun Coal Mine, the coal deposit under the railway special line reaches 35 million tons. It is particularly important to ensure the safe operation of the railway while promoting the comprehensive development and utilization of coal resources. In order to accurately predict the dynamic process of mining surface movement and deformation，and provides the foundation for the measures of surface mining resistance and the optimization design of working face，based on Weibull time function model. Firstly, by combining theoretical calculation with software processing, the variation characteristics of the function in terms of subsidence amount, subsidence velocity and subsidence acceleration are analyzed. The results are consistent with the surface subsidence characteristics of Changcun Coal Mine, indicating that the function has good completeness in time and space. Then on the basis of collating and analyzing the monitoring data of surface dynamic subsidence in Changcun Coal Mine, selected the 11 monitoring data fitting processing and comparison analysis, the research determines the main influencing factor of the model parameters c and k, at the same time put forward the parameters c and k change rule and change curve, and through the measured data demonstrates the reliability. Finally, based on the above research, we select No. 22 and No. 30 monitoring point, and through dynamic comparison and analysis of measured and predicted values of subsidence, it is found that the predicted relative accuracy is 2.11% and 2.66%, respectively. Weibull time function model and parameter determination method have certain theoretical and practical significance for improving the prediction accuracy of mining surface dynamic subsidence process.