Abstract: According to the shortage of Knothe time function of the sinking speed in describing the dynamic subsidence process,the improved two-parameter Knothe time function was adopted to establish the dynamic subsidence model,where the decision coefficient c of overburden rocks and exponent k value were resolved by using the least square method,and the maximum subsidence of W0 was determined through the surface movement observation station by fitting the measured data;R2 coefficient evaluation precision was adopted and applied in a mine in Huainan and in 1242 (1) working surface movement observation station data model to verify the accuracy of the model.The maximum subsidence point,MS29 and ML44 in the fitting decision coefficient of each observation period were 0.983 6 and 0.975 7.The fitting decision coefficients of the observation value and the expected value of each observation point in dip and strike observation line were 0.995 3 and 0.958 2 separately as the advancing is half (328 d).The calculation results show that the dynamic prediction of two-parameter Knothe time function model on the whole process of mining subsidence of 1242 (1) working face is accurate and reliable.

Key words: Mining subsidence, Dynamic prediction, Two-parameter Knothe time function, Model parameter, Least square method