A doublegeneticalgorithmfortheMRCPSP/max

This paperpresentsaheuristicsolutionprocedureforaverygeneralresource-constrainedproject schedulingproblem.Here,multipleexecutionmodesareavailablefortheindividualactivitiesofthe project. Inaddition,minimumaswellasmaximumtimelagsbetweendifferentactivitiesmaybegiven. The objectiveistodetermineamodeandastarttimeforeachactivitysuchthatthetemporaland resource constraintsaremetandtheprojectdurationisminimised.Projectschedulingproblemsofthis type occure.g.inprocessindustries.Theheuristicisatwo-phasedgeneticalgorithmwithdifferent representation,fitness,crossoveroperator,etc.,ineachofthem.Oneofthecontributionsofthepaperis the optimisationinthefirstphaseofaproblemdualtotheoriginal,thesearchingforthebestmodesof the activities.Computationalresultsshowthatthealgorithmoutperformsthestate-of-the-art algorithms inmediumandlargeinstances.