Journal of Heuristics

, Volume 5, Issue 4, pp 419–436

Best Practice Simulated Annealing for the Airline Crew Scheduling Problem

  • Thomas Emden-Weinert
  • Mark Proksch

DOI: 10.1023/A:1009632422509

Cite this article as:
Emden-Weinert, T. & Proksch, M. Journal of Heuristics (1999) 5: 419. doi:10.1023/A:1009632422509


We report about a study of a simulated annealing algorithm for the airline crew pairing problem based on a run-cutting formulation. Computational results are reported for some real-world short- to medium-haul test problems with up to 4600 flights per month. Furthermore we find that run time can be saved and solution quality can be improved by using a problem specific initial solution, by relaxing constraints “as far as possible”, by combining simulated annealing with a problem specific local improvement heuristic and by multiple independent runs.

airline crew scheduling simulated annealing pairing problem 

Copyright information

© Kluwer Academic Publishers 1999

Authors and Affiliations

  • Thomas Emden-Weinert
    • 1
  • Mark Proksch
    • 1
  1. 1.Institut für InformatikHumboldt-Universität zu Berlin, December, 14th, 1998Germany

Personalised recommendations