A problem of aggregate production planning, with uncertainty about the fluctuation in demand, is formulated by a stochastic optimization model, with quadratic criterion and linear constraints. Difficulties in finding a global optimum solution to the problem lead to the proposal of an adaptive approach which is easy to implement computationally and which is based on the formulation of a deterministic equivalent problem, the solution for which is periodically reviewed through a classical procedure of literature. An example, where the inventory balance system is subject to weak and strong variability in actual demand, is employed to analyze the behavior of the proposed approach. Finally, the results provided by the proposed approach are compared with another suboptimal approach, the main characteristic of which is not allowing periodic reviews.
Inventory control; planning; optimization; stochastic process; simulation