Modelling the load levelling problem in master production scheduling for MRP systems

When infeasible master production schedules (MPS) are detected during the rough-cut capacity planning phase, material requirements planning (MRP) systems rely on the user's expertise to solve this problem. The usual procedure is to shift production either earlier or later in the planning horizon or to increase the capacity at some work centres during given periods of time. The main challenge is to identify what the best changes are in order to maintain a reasonable load profile across all work centres. In this paper we develop optimization models to support the scheduler's task of resolving capacity infeasibilities. The models take into consideration a variety of costs and manufacturing characteristics to provide the scheduler with maximum flexibility. In an extension to the basic model, the scheduler is also able to control the degree of difference between a proposed schedule and the original MPS.