SINGLE-STAGE MULTI-PRODUCT PRODUCTION AND INVENTORY SYSTEMS: AN ITERATIVE ALGORITHM BASED ON DYNAMIC SCHEDULING AND FIXED PITCH PRODUCTION

The paper presents a model and an algorithm for solving a class of stochastic economic lot sizing problems (SELSP). SELSP is important because it is very common in both manufacturing and process industries. The algorithm is related to the manufacturing strategy which consists of dynamic scheduling and fixed pitch production. These strategies are rather unexplored in the literature; yet, they are promising since they give flexibility to the solution. The algorithm returns a near-optimal solution composed of all the parameters needed to operate the production system. In order to do so, it iteratively combines two procedures: (i) a method of successive approximations which chooses progressively better values for the production pitch and the lot sizes, and (ii) a discrete stochastic simulation routine to find the order points. The algorithm proved to be fast enough to allow its frequent use in real world problems, on computers in current use.