Abstract:
This study considers the single machine due date assignment problem involving dynamic arrivals and family setups. A rolling horizon approach following periodic rescheduling strategy is developed to quote due dates. At speci c time points, jobs in the system are rescheduled, and due dates of newly arrived jobs are quoted taking the constructed schedule into account. According to the proposed policy based on this approach, a static scheduling problem is solved periodically to determine the order of processing and the due dates of jobs in dynamic environments. For this static problem two separate mixed integer linear programming formulations are developed. Using one of the formulations two experiments are conducted. In the rst experiment, the performance of the solver is tested by solving a set of various-sized static scheduling problems generated with di erent parameters. The environmental parameters and the solver parameters of the second experiment are set based on the results of the rst experiment. In the second experiment, a set of policies is constructed using di erent parameters, and their performances are analyzed in distinct dynamic production environments. The changes in the system performance measures with respect to the parameters of the proposed policy are examined for each environment.
