<p>As a current research focus in the field of rail transportation, virtual coupling operation is an important means to enhance transport capacity and scheduling flexibility. In light of the disturbances and safety challenges faced by train control system in real environments, this paper proposes a model predictive control approach with a variable adaptive zone. An adaptive zone track occupancy control strategy is proposed based on the train operation mode, aiming to dynamically adjust the operating intervals of virtual coupling. This strategy is integrated with a fuzzy logic system to quantify disturbance signals, thereby enabling a dynamic adjustment mechanism for the adaptive zone. Furthermore, a predictive feedback mechanism is introduced into the virtual coupling controller. This mechanism adjusts the target operating curve of the train set based on the real-time state of the trains and the control strategy for the adaptive zone. To validate the effectiveness of the proposed method, simulation analysis is conducted using actual track data and train dynamics data generated by the onboard train controllers. The results demonstrate that, after adopting the adaptive cooperative controller, the average amplitude of the following train is significantly reduced, indicating effective attenuation of disturbances and a notable improvement in system stability. However, under maximum amplitude conditions, larger fluctuations occur at certain moments due to the cumulative effect of speed errors in the control strategy. Nevertheless, the dynamic adjustment mechanism of the adaptive zone effectively suppresses these fluctuations and restores system stability.</p>