Abstract: There are three groups of faults developed in Huashugou copper mine of Jingtieshan.The rock mass structure of the mine is complete and the rock strength is high.Affected by the three groups of faults, the mine engineering geological problems are easy to occur in the mining process, and the roof fall disasters occur frequently in the roadway.The mine production safety is greatly threatened.Therefore, it is of great practical significance to deploy the microseismic monitoring system in Huashugou mining area of Jingtieshan and carry out the research on the intelligent data processing and analysis of microseismic data based on deep learning technology.According to the actual production situation of the mine, the mining area is divided into five monitoring zones by K-means clustering analysis algorithm.At the same time, a set of effective and reliable disaster early warning method based on the daily frequency of microseismic events, the cumulative number of microseismic events, distribution of βn and CUFIT model and other early warning parameters and methods of cumulative energy release change is formed.The study results have successfully given early warning to the ground pressure disaster events in monitoring fourth division of the mining area.