In the present work, targeting at the perfect coupled tasks of two adjacent high‐speed trains under the virtual coupling tracking processes, a novel prescribed performance tracking control method is developed, where both the system uncertainties and the actuator constraints are taken into account. Notice that by utilizing a novel error conversion method, the singularity problem may occur in the prescribed performance control could be avoided, while the input constraints are also satisfied by using a smooth function to approximate the non‐smooth input saturation function. The system uncertainties are estimated by adopting the radial basis function neural network. Different with existing works, the initial speeds of the adjacent trains can be nonidentical. Furthermore, the stability of the proposed controller is rigorously analyzed under the framework of Lyapunov theory. To demonstrate the effectiveness of the proposed control approach, three different uncoupled cases are considered, where the corresponding tracking processes from coupling to completely coupled are described in details.