Abstract: Gas accidents occur frequently in the process of coal mining, and the structural area is a high-risk area of gas accidents. Therefore, accurate prediction of hidden structures is the key to effective prevention and control of coal mine gas outburst. In order to realize the effective identification of concealed structures, this study takes the No.9 coal seam of Longfeng Coal Mine in Lindong, Jinsha County as an example, and comprehensively uses field monitoring and laboratory experiments to study. The system collects the dynamic data of gas volume fraction and emission before and after passing through concealed structures in typical working faces such as 120905 transport roadway and 120910 air roadway. Coal samples with different distances (20 m, 50 m, 100 m) from known faults were collected. The mechanical-seepage coupling tests under different confining pressures (1-2 MPa), different gas pressures (1-2 MPa) and different loading rates (40-80 N / s) were carried out by using TAW-2000 rock triaxial testing machine. The strength, deformation and permeability evolution characteristics of coal samples were quantitatively analyzed. The results show that the closer the coal body is to the fault, the more broken the structure is, the lower the strength is, and the higher the permeability is. The permeability of the coal body at 20 m from the fault is about twice that at 100 m. Under all stress paths, the permeability of coal body shows an evolution trend of decreasing first and then increasing sharply, and the initial, minimum and failure permeability of coal samples at 20 m from the fault is 2 times that of coal samples at 100 m, which proves that the near-fault coal body has stronger gas migration ability. On-site monitoring found that when the gas volume fraction or emission is significantly abnormal, it is often indicated that there is an undisclosed structure in front of it. Based on this, a comprehensive identification method of “gas abnormal response-permeability evolution identification” was constructed and successfully applied in the heading face, which improved the accuracy of structural prediction and the ability of gas disaster prevention and control.