Reinforced Prescribed Performance Control for Virtually-Coupled Trains With Saltatory Targets and Switching Constraints

Virtual Coupling (VC) has emerged as a promising strategy for railway system control, offering significant enhancements to rail-line capacity and addressing uneven transport demand by dynamically minimizing the headway distance between trains when necessary. Consequently, the operating targets of VC rear trains, which can be influenced by operational environments and railway line conditions, must be promptly adjusted based on the spacing from the adjacent leading train. This adjustment ensures operational safety within the train movement authority. Different from traditional continuous prescribed performance control (PPC), this paper proposes a discontinuous reinforced prescribed performance control (RPPC) strategy, which takes into account both actual and converted errors and introduces innovative segmented dynamic control strategies designed for continuous operation time periods and discontinuous impulse points. To prevent unnecessary disruptive switching during transient moments, switching constraints (SCs), based on the speed and position of the rear train, are established. Additionally, leveraging Lyapunov theory, the stability of the closed-loop system for train operation is rigorously proven. Finally, a simulation scenario for a 3-train formation, along with a further data-based semi-physical simulation of a 2-train system, are presented for demonstrating the feasibility of the main results.