A Utilização do bootstrap em um sistema de estoque multi-estágio

Abstract

An accurate forecast of demand and good management of inventories of items MRO (Maintenance, Repair and Operation) are essential for companies due to high inventory costs associated with these items. Several authors in the literature have developed methods of forecasting demand and discussed which statistical distribution best suited for modeling the demands of these items in a multi-echelon system. However, the bootstrap technique with excellent performance in single-echelon systems has been little used in more complex systems. Motivated by this information, this paper proposed the insertion of the bootstrap technique in a discrete simulation to model the demands of repair items on a multi-echelon system. As shown by the results, the bootstrap presented behavior similar as showed when it is tested in single-echelon systems.