Estimation of Inventory Item Demand Distributions: Modeling Item Migration at the Defense Electronics Supply Center

摘要

The objective of this thesis is to stimulate the migration of items between management categories in a large inventory system. The approach taken is to model demand from distributions created by grouping items rather than from individual demand distributions. Items are grouped according to characteristics such as demand frequency, average requisition size, and price. An extensive historical database is used to develop the simulation input distributions and to compare simulation results against actual inventory system demand and migration figures. The empirical requisition size and daily demand distributions exhibit non-random tendencies and extreme values which cannot be modeled using common theoretical distributions. Simulated demand and migration is dependent on the characteristics used to define the item groups. Simulated item migration occurs, but is not representative of the actual migration present with the item sample. Keywords: Inventory analysis; Statistical analysis; Statistical distributions; Data acquisition; FORTRAN; and Computer programs.