Abstract: Aiming at the big data analysis of hydraulic support resistance monitoring of fully mechanized mining face, the prediction and early warning of incoming pressure and the influencing factors of incoming pressure, a mining pressure prediction model based on big data is proposed. According to the FLAC3D numerical simulation and the actual monitoring results of Liuzhuang Coal Mine , divide the push mining into three stages: stable push mining, shutdown maintenance, and rapid push mining. The following conclusions are drawn: the prediction accuracy of the incoming pressure value of big data is about 65.5%, and the prediction accuracy rate of the incoming pressure range is about 51.7%. ; The larger the mining distance in a single working day, the greater the subsidence of the roof, and the greater the pressure value of the support; in the stable mining stage of the working face, the number of times of pressure is less, and the step distance of the pressure is large; In the rapid mining stage, the pressure is frequent and the pressure step is small. According to the change of the pressure in the corresponding stage of mining, the predicted correction value K of Liuzhuang Mine is ±1.5.