Quantum-Inspired Evolutionary Neighborhood Search for Railway Arrival-Departure Track Utilization under Short-Term Disturbances

This paper models railway-station recovery by representing arrival, track-occupancy, and departure operations as zone-level resource-occupation intervals. It jointly considers train delays and resource reassignment costs, then proposes a quantum-inspired evolutionary algorithm with neighborhood search (QEA-NS). On perturbation instances derived from GTFS timetable data for Frankfurt Hauptbahnhof, QEA-NS produced 388 total delay minutes versus 519 for CP-SAT, a 25.2% reduction. Across 10 random instances, it achieved lower total delay in every case, with mean delay of 390.5 minutes versus 673.8 minutes for CP-SAT. The trade-off is longer solution time.