In operational production planning several companies have to size and schedule production lots on a set of parallel machines to satisfy forecasted demand, while facing sequence dependent changeover times and costs. Motivated by a case study in a beverage company, we exploit the practice of rolling basis planning to develop efficient approaches to the problem. The horizon is decomposed in two parts: the initial periods explicitly consider the production sequences to obtain detail schedules, while in the remaining periods a rough plan is generated to give an estimation of future costs and capacity. Several modeling alternatives proposed in the literature are reviewed and anew formulation that includes the setup loss in the future periods based on the loss experienced in the detail part of the horizon is proposed. An important contribution is an innovative iterative method to improve the accuracy of the approximate parameters used in the context of the simplified models. We assess the performance of several alternatives by simulating the implementation of solutions on a rolling horizon by using instances generated based on the features arising in the beverage industry. The tests simulate the planning environment in practice by generating demand forecasts followed by the implementation of the production plans. The results show that applying the iterative approach on the approximate methods can generate solutions that better trade-off costs and stockouts.
|Original language||English (US)|
|Number of pages||27|
|State||Published - 2015|