Abstract
This study is motivated by the production management problem found in many large-volume MTO systems. In such an environment, a decision maker has to evaluate the requested due dates if are capacity feasibility, and further, to determine how to distribute each confirmed order with arbitrary ready date and due date to parallel machines so that the demand quantity and the due date can be met. Sequence- and machine-dependent setup times in unrelated parallel machine systems are considered in this study. In order to finish promised orders on time, splitting the production items into parallel machines for simultaneous processing is necessary. Two different approaches are proposed for the two hierarchical levels of production management – the preemptive earliest due date (PEDD) CRP approach in order entry level and the lexicographic goal programming approach in operational scheduling level. Two priority objectives are considered in the goal programming, the first priority is to minimize the total tardiness and the second priority is to minimize the total setup time. Therefore, two different mixed integer programming (MIP) formulations are proposed for the two priority objectives. Furthermore, two models with two different assumptions on job splitting are investigated in the operational scheduling level. Model 1 allows at most one setup for a job on a machine, while Model 2 allows multiple setups. Because of the assumption difference, the two models adopted two different modeling techniques of binary variables. The immediate-precedence variables modeling technique is used in Model 1; whereas, the sequence position variable modeling technique is used in Model 2. The experimental result shows that the proposed approach can effectively and efficiently solves the problems.