A multi-objectiveapproachforrobustairlinescheduling

We presentamemeticapproachformulti-objectiveimprovementofrobustnessinfluencingfeatures (called robustness objectives) inairlineschedules.Improvementoftheobjectivesisobtainedbysimulta- neous flightretimingandaircraftrerouting,subjecttoafixedfleetassignment.Approximationsofthe Pareto optimalfrontareobtainedbyapplyinga multi-meme memeticalgorithm. Weinvestigate biased meme selection to encourageexplorationoftheboundariesofthesearchspaceandcompareitwith random memeselection. Anexternalpopulationofhighqualitysolutionsismaintainedusingtheadaptive grid archivingalgorithm. The presentedapproachisappliedtoinvestigatesimultaneousimprovementof reliability and flexibility in realworldschedulesfromKLMRoyalDutchAirlines.Experimentalresultsshowthattheapproach enables ustoobtainscheduleswithsignificantimprovementsfortheconsideredobjectives.Alargescale simulation studywasundertakentoquantifytheinfluenceoftherobustnessobjectivesontheoperational performance oftheschedules.Rigoroussensitivityanalysisoftheresultsshowsthattheinfluenceofthe schedule reliabilityisdominantandthatincreasedscheduleflexibilitycouldimprovetheoperational performance.