Abstract: In order to study the vibration characteristics of shearer’s spiral drum under various occurrence conditions, the coal-rock contact model was optimized by taking the MG2×55/250–BWD thin seam shearer as the engineering object, and discrete element model of the coal wall under various cutting conditions similar to the actual occurrence conditions were established. Combined with DEM–MFBD (Discrete Element Method-Multi Flexible Body Dynamics) two-way coupling numerical simulation method, the two-way coupling simulation platform of the rigid-flexible coupling virtual prototype model of the shearer cutting section and the discrete element model of the coal wall was built. Through simulation, the cutting process of the spiral drum under different coal wall working conditions was obtained, and the variation law of the vibration characteristics of the spiral drum under different occurrence conditions was analyzed. The results show that: during the cutting process, the spiral drum vibrates to different degrees in three directions, of which the vibration acceleration in the cutting resistance direction is the largest, the vibration acceleration in the traction resistance direction is the second, and the vibration acceleration in the lateral force direction is the smallest. With the increase of the hardness of the gangue and the proportion of the number of layers in the model, the vibration intensity of the spiral drum during the cutting process continues to increase, and the difference between the effective values of the maximum vibration acceleration reaches 4 403.149 mm/s². The short-time Fourier transform was used to convert one-dimensional vibration signal into two-dimensional time-frequency spectrum image, and the characteristics of vibration information change under different working conditions were well preserved in time-frequency domain. The feature sample effect of time-frequency spectrum image is better than that of time-domain one-dimensional signal curve under various working conditions. The information such as the location, scope and shape of feature cluster of dominant frequency energy have obvious differences. Even if the number of rock parting is different, the distribution form of energy characteristics in the time-frequency spectrum image is also significantly different under complex working conditions where there are differences in the firmness coefficient of the rock parting. Through vibration modal analysis, it is found that with the increase of the hardness of the gangue in the coal wall, the deformation of each part changes, and the change of the pick part is the strongest. Building a vibration signal testing experimental platform for shearer based on similarity theory. The cutting process of spiral drum under different working conditions was tested experimentally. By tracking the vibration state of the spiral drum, it is found that the vibration change law is consistent with the results of the two-way coupling numerical simulation. The error between the effective value of the vibration acceleration of the spiral drum obtained by the DEM–MFBD numerical simulation method and the experimental data inferred based on similarity ratio is smaller than the error between the DEM discrete element numerical simulation method and the experimental data, which verifies the accuracy of the DEM–MFBD numerical simulation method. The research results are of great significance for the working reliability of the lifting spiral drum, and also provide a new method for obtaining data signals during the construction of the coal and rock cutting state recognition system for "unmanned" intelligent mining.