Airport Service Vehicle Scheduling

Kurzfassung

Airport service vehicles, such as luggage trailers and passenger buses, service an aircraft after the aircraft arrives and before it departs. The timely arrivals of these vehicles help ensure efficient use of airport resources. This research investigates algorithms for airport service vehicles. A mixed integer linear program is proposed, minimizing service provider fuel costs and air carrier delays. The formulation of the integer programming problem is modified to aid solution search strategies. A genetic algorithm heuristic borrowed from aircraft arrival scheduling is introduced for finding approximate solutions relatively quickly, in addition to an exact solution method making use of branch and bound techniques specially designed for this problem. The various algorithms are tested using simulations of service provider dispatch problems at Hamburg and Dallas-Fort Worth Airports. Results show that plan based service vehicle scheduling reduces both delay and fuel costs over passive strategies, often 20% or more. The genetic algorithm based heuristic also reduced delay and fuel costs while incurring computational burdens significantly below those of the optimal search strategy.