Cooperative decision-making and control strategies for unsignalised intersections with emergency vehicle priority

Emergency vehicles (EVs) often experience delays at congested intersections, undermining time-critical rescue operations. Enabled by connected and automated vehicle (CAV) technologies and Vehicle-to-Everything (V2X) communication, intersections can be managed cooperatively, but EV priority that balances EV efficiency and surrounding traffic performance remains underexplored. This paper proposes MILP-LC, a mixed-integer linear programming (MILP) approach integrating a lane-changing yielding strategy. Vehicles ahead of the EV are instructed to proactively change lanes to create space, while the MILP model optimises vehicle passing orders to minimise disruption to surrounding traffic. A level-based control algorithm executes the plan to ensure orderly movement through the intersection. Simulation results show that, compared with a heuristic depth-first spanning tree algorithm and conventional EV signal preemption, MILP-LC significantly reduces EV travel time while maintaining overall traffic efficiency with only a slight increase in fuel consumption. Benefits are more pronounced under high traffic volumes and short time headways.