The Logarithmic Poisson Gamma Distribution: A Model for Leadtime Demand.
Abstract
Many of the inventory models which are used in practice rely upon knowning the probability distribution of demand over a leadtime. The common assumption is that this distribution is normal. However, in certain circumstances, the normality assumption may be inappropriate. We consider a process in which demand occurrences obey a stationary Poisson process and at each occurrence a random number of units are demanded. Furthermore, the leadtime for replenishment is assumed to be random as well. Assuming a logarithmic compounding distribution and a gamma distribution of leadtime we derive the exact distribution of total demand in a leadtime. We call this the Logarithmic-Poisson-Gamma (LPG) distribution. An approximation is derived which is equivalent to a scaled version of the negative binomial distribution. In the final section we derive the first four central moments.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1981
- Accession Number
- ADA124540
Entities
People
- Steven Nahmias
- W. Steven Demmy