From State-of-the-Art Static Fleet Assignment to Flexible Stochastic Planning of the Future

  • Authors:

  • Sven Grothklags;Ulf Lorenz;Burkhard Monien

  • Affiliations:

  • Lufthansa Systems Berlin GmbH Airline Planning & Control, Berlin D-10587;Institute of Mathematics, Darmstadt University of Technology, Darmstadt, D-64289;Faculty of Computer Science, Electrical Engineering and Mathematics, University of Paderborn, Paderborn, D-33102

  • Venue:
  • Algorithmics of Large and Complex Networks
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Given a flight schedule, which is a set of non-stop flights, called legs, with specified departure and arrival times, and a set of aircraft types, called subfleets, the fleet assignment problem is to determine which aircraft type should fly each leg. The objective is to maximize the overall profit. Usually, planners assume that precise input data are determininstically available at planning time. As a consequence, an important insufficiency of modern industrial plans, especially flight-plans, is their lack of robustness. Disruptions prevent from operating as planned before and induce high costs for trouble shooting. One reason for the uncertainties in the data for fleet assignment is that this problem is part of a long row of optimization problems of an airline. Therefore, important restrictions of later steps like connection dependent ground times should be considered in the fleet assignment problem. We show how connection dependent ground times can be added to the fleet assignment problem and presents three optimization methods, varying in run time and solution quality, that can solve real-world problem instances with more than 6000 legs within minutes. Moreover, real or believed non-determinism leads to inevitable uncertainties in input data. As a consequence, instead of a traditional plan, a flexible strategy which reacts on different relizations of the uncertain data is demanded. The Repair Game is a formalization of a planning task, and playing it performs disruption management and generates robust plans with the help of game tree search. We introduce the game and present experimental results of a feasibility study.