Abstract: As a key part of virtual coupling system, rail transit vehicle-to-vehicle communication has a profound impact on the coordinated operation of virtually coupled train formation. Based on the analysis of the probability distribution characteristics of communication delays, this study explores the impact of communication delays on the virtual coupling of trains, providing theoretical support for the design of vehicle-to-vehicle communication system performance. The article first proposes a train dynamics model and uses model predictive control (MPC) to achieve coordinated operation control between multiple trains in the virtual coupling. Through simulation experiments with different probability distribution scenarios of communication delays, the study analyzes the impact of communication delays on the virtual coupling control system for trains. The experimental results show that larger communication delay disturbances exacerbate oscillations in the virtual coupling system. These issues pose challenges to the smooth operation and energy efficiency of trains.