An Evolutionary Algorithm for Large Scale Set Covering Problems with Application to Airline Crew Scheduling

Abstract

The set covering problem is a paradigmatic NP-hard combinatorial optimization problem which is used as model in relevant applications, in particular crew scheduling in airline and mass-transit companies. This paper is concerned with the approximated solution of large scale set covering problems arising from crew scheduling in airline companies. We propose an adaptive heuristic-based evolutionary algorithm whose main ingredient is amechanism for selecting a small core subproblem which is dynamically updated during the execution. This mechanism allows the algorithm to find covers of good quality in rather short time. Experiments conducted on real-world benchmark instances from crew scheduling in airline companies yield results which are competitive with those obtained by other commercial/academic systems, indicating the effectiveness of our approach for dealing with large scale set covering problems.