Premium
Optimization methods for large‐scale multiechelon stockage problems
Author(s) -
Leblanc Larry J.
Publication year - 1987
Publication title -
naval research logistics (nrl)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.665
H-Index - 68
eISSN - 1520-6750
pISSN - 0894-069X
DOI - 10.1002/1520-6750(198704)34:2<239::aid-nav3220340209>3.0.co;2-t
Subject(s) - computer science , bounding overwatch , lagrangian relaxation , mathematical optimization , operations research , service (business) , simplex , mathematics , artificial intelligence , geometry , economy , economics
I examine the problem of determining inventory stockage levels and locations of different parts in a multiechelon system. This stockage problem is complicated by parts commonality—each part may be used by several different end items. Stockage levels and locations of each part affect the availability of end items that use the part, since an end item will be out of service if it requires a part that is not available. Of course, if the part is available at another nearby location, then the end item will be out of service for a shorter period of time. An essential feature of any model for this problem is constraints on operational availability of the end items. Because these constraints would involve nonconvex functions if the stockage levels were allowed to vary continuously, I formulate a 0–1 linear optimization model of the stockage problem. In this model, each part can be stocked at any of a number of prespecified levels at each echelon. The model is to minimize stockage cost of the selected items subject to the end‐item availability constraints and limits on the total weight, volume, and number of different parts stocked at each echelon. Advantages and disadvantages of different Lagrangian relaxations and the simplex method with generalized upper‐bounding capability are discussed for solving this stockage model.