Peter Berling and Victor Martinez de Albeniz ()
Additional contact information
Peter Berling: Lund University
Victor Martinez de Albeniz: IESE Business School, Postal: Research Division, Av Pearson 21, 08034 Barcelona, SPAIN
Abstract: In this paper, we present a continuous model to optimize multi-echelon inventory management decisions under stochastic demand. Observing that in such continuous system it is never optimal to let orders cross, we decompose the general problem into a set of single-unit sub-problems that can be solved in a sequential fashion. When shipping and inventory holding costs are linear in the stage, we show that it is optimal to move the unit associated with the k-th next customer if and only if the inventory unit is held in an echelon located within a given interval. This optimal policy can be interpreted as an echelon base-stock policy such that the base-stock is initially increasing and then decreasing in the stage. We also characterize the optimal policy when costs are piecewise-constant. Finally, we study the sensitivity of the optimal base-stock levels to the cost structures.
Keywords: multi-echelon; optimal control; unit-tracking decomposition
35 pages, February 1, 2011
Full text files
DI-0905-E.pdf
Questions (including download problems) about the papers in this series should be directed to Noelia Romero ()
Report other problems with accessing this service to Sune Karlsson ().
RePEc:ebg:iesewp:d-0905This page generated on 2024-09-13 22:20:02.