Integrated Routing and Controlled-Segment Scheduling in Corridor-Based Drone Logistics Systems Under Minimum Headway Constraints

Abstract

Predefined low-altitude corridors create a coupled routing–scheduling problem when multiple drone routes enter the same controlled segment. This study separates an upstream control hub from its scarce directed hub–segment resource and develops an event-expanded continuous-time mixed-integer linear programming model with optional fleet activation, complete-route energy and capacity checks, release precedence, minimum entry headway, holding, and downstream delay propagation. A headway-aware large neighborhood search (HA-LNS) combines route neighborhoods with a finite serial event decoder. Gurobi proves optimality on three small instances, and fixed-route timing MILPs exactly match the decoder, including for a repeated physical-hub visit. Across ten matched networks per scale, HA-LNS changes the mean objective relative to route-only LNS by 0.01%, 0.90%, and 2.33% at nominal scales 30, 50, and 100. Under high conflict-resource density, the reduction reaches 5.77%, while mean holding falls from 2.054 to 0.025 min. Simulated annealing is 1.04% better at scale 50 and statistically indistinguishable at scales 30 and 100, showing that the contribution is conflict-aware integration rather than universal heuristic dominance. The framework identifies directed-resource density as the main condition under which temporal coordination materially improves route decisions.