Research Project

STOCHASTIC CAPACITATED LOT SIZING WITH MULTIPLE DECISION STAGES

Exemplary production plan determined with the multi-stage approach

Stochastic capacitated lot sizing aims at determining optimal production quantities for products to be produced without interruption on a single machine, when a setup procedure is necessary to prepare for the production, which is costly and/or takes some time and thereby reduces the capacity available for production. The objective is the minimization of total costs for setup activities, inventory holding and overcapacity. For practical applications it is crucial to consider several parameters as uncertain. In many cases the future demand is not known with certainty. In those cases the demand is predicted using forecasting methods to determine the parameters of the distribution of expected demand. In those cases a planning algorithm which allows some flexibility to adjust the production plans when unexpected demand realizations are observed is beneficial, while also a certain amount of stability is necessary to avoid nervousness in the production system. Another relevant aspect for practical application is the reduction of variability of the distribution of the objective function value, which can be achieved by applying methods from risk averse optimization.

Field of Research