Solving the Integrated Schedule Generation and Fleet Assignment Problem: an ACOBased Metaheuristic Approach

Daniel J. Caetano, Nicolau D. F. Gualda
2015 Journal of Transport Literature  
Article Info Abstract Keywords: air transportation schedule generation fleet assignment metaheuristic ant system Traditionally, the initial steps on airline planning -Schedule Generation and Fleet Assignment problemsare solved separately. This traditional approach usually leads to suboptimal solutions, since flight profitability -the decision criteria to schedule a flight -depends on what aircraft type will be used on that flight. On the other hand, the type of aircraft assigned to a flight
more » ... ned to a flight will be different accordingly to the available scheduled flights. Because of this interdependence, airlines avoid complete redesign of their flight network, adopting a conservative approach and slightly improving existing suboptimal schedules over time. The integrated solution for both problems, albeit desirable, leads to large-scale models of the NP-Hard class. Some of the original linear constraints may become non-linear in the integrated problem, bringing further complexity to the solution process. This article presents a linear programming formulation of this integrated problem along with a heuristic approach, called MAGS, based on the ACO metaheuristic. Both the exact solution and the one provided by MAGS are obtained and compared for the case of a Brazilian airline. The results show the applicability of MAGS to real world cases, presenting solutions with objective function values distant no more than 6% of the optimum and much lower processing times than LP model on more complex configurations.
doi:10.1590/2238-1031.jtl.v9n3a6 fatcat:ojh7qmkgn5dnpfecga5xjf5v3u