In order to solve the problem of intelligent identification of coal-rock interface of shearer in fully mechanized coal mining face, a CPS-based coal-rock identification system structure of shearer is proposed. A hybrid system simulation design was carried out for the system, a wireless sensor network system model based on Ptolemy II was constructed, and the design and modeling of the wireless sensor node was completed. The overall simulation of the system was carried out, the experimental data was extracted from the physical prototype system, and the information processing system model was constructed based on the wavelet packet feature extraction and the PSO-BP neural network algorithm. Through the joint simulation of Ptolemy II and Matlab, the reliability of the CPS-based coal-rock identification system is verified.The simulation results show that the average error value of the system co-simulation is larger than the result obtained by the single method simulation, the maximum error is 0.027 924, and the error is still within the allowable range. The results obtained through the simulation of the hybrid system more truly reflect the coal-rock identification process of the shearer, and the prediction accuracy is higher. The research provides a certain contribution to the modeling and simulation of the CPS hybrid system.