Train movement high-level model for real-time safety justification and train scheduling based on model predictive control

In the infrastructure of networked operation of high-speed trains, the justification of operation safety and the decision of scheduling strategies should be undertaken in a real-time way, which depends on the credible prediction model of train movements. This paper will incorporate the train movement high-level model into the real-time conflict detection and train scheduling based on the principle of model predictive control (MPC). The proposed model describes the restrictive, synergistic and autonomous train movements with continuous accelerations and decelerations in the discrete time and continuous space. The sufficient modeling accuracy can be achieved if the adjustable parameters are properly configured for the MPC-based control and management of train operations. Consequently, the detection of block-section occupation conflicts considering the virtual junctions and the decision of feasible scheduling strategies possess the considerable confidence level and safety guaranty. The conflict detection and the operation optimization are implemented over the rolling horizon according to the real-time feedback information. The numerical results demonstrate the utility and rationality of the proposed model for the real-time safety justification and the scheduling strategy evaluation.