A simple and efficient metaheuristic for the dynamic flight scheduling problem

One of the biggest problems in the airline industry is facing the uncertainty created by passengers demand. In this work, we introduce a metaheuristic that uses updated demand forecasts to optimise a given flight schedule. The algorithm uses two main techniques to solve the problem: (i) re-timing, aimed at changing the scheduled time of a flight by delaying or advancing it to allow new connections; and, (ii) re-fleeting is focused on swapping aircraft between flights to increase aicraft’s occupancy and better adjust it to demand. The algorithm’s objective is to maximise airlines revenue by increasing the occupancy of each flight. This metaheuristic has been tested in different scenarios of up to 600 flights with promising results in low CPU times.