A large neighbourhood search approach to airline schedule disruption recovery problem

Abstract

The occurrence of unplanned aircraft shortages and disruption of flight schedules during the day-to-day operations of airlines is inevitable. When equipment failure causes unsafe flight, the aircraft will be grounded or temporarily delayed when the weather shuts down the airport or the required flight crew is unavailable. Real-time decisions must be made to reduce revenue loss, passenger inconvenience and operating costs by reallocating available aircraft and cancelling or delaying flights. A large neighbourhood search algorithm is used in this research to construct a feasible and efficient solution to the airline schedule disruption recovery problem. We aim to reduce the aircraft turn-around times, including total delay time, the number of flight adjustments and the number of flights delayed for more than one hour, as an objective function. Ten real-life cases are solved, and the proposed approach yields an approximate 50% improvement in solution quality.