STOCHASTIC INVENTORY CONTROL AND THE ASSUMPTION OF NON-INTERCHANGEABILITY (VARIABLE LEAD-TIME).
Item
-
Title
-
STOCHASTIC INVENTORY CONTROL AND THE ASSUMPTION OF NON-INTERCHANGEABILITY (VARIABLE LEAD-TIME).
-
Identifier
-
AAI8611370
-
identifier
-
8611370
-
Creator
-
NASRI, FARROKH.
-
Contributor
-
Georghios P. Sphicas
-
Date
-
1986
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Business Administration, Management
-
Abstract
-
In less than half a century, literally thousands of articles have been published in the area of inventory management, each with its own set of objectives and assumptions. This research is another attempt to expand the body of knowledge in this field. The significance of this research is two-fold: (a) its application to some specific production problems and (b) its motivation of further studies investigating the significance of differences between inventory models that allow "cross over" and those that simply avoid it.;This dissertation deals with a (Q,R) inventory model with deterministic demand and finite stochastic leadtimes. In contrast to the mainstream of inventory literature, we allow for crossing of orders and make a more restrictive assumption--i.e., non-interchangeability (non-substitutability) of orders.;It is shown that, in general, the possibility of cross over arises if and only if the range of lead-time is larger than the optimal cycle time.;A stochastic generalization is given for basic EOQ's total optimal cost function when backorders are allowed. Some additional extensions are obtained for the case of uniformly distributed lead-times. After extensive algebraic manipulation, simplified cost expressions are obtained for other ranges of the model.;Lower and upper bounds of optimal costs are obtained for each model. Also, the shape and properties of the optimal cost as a functin of the range of lead-time is given.;Finally, we obtain the probability of cross over of two consequent orders. We also extend the results to the case of uniform lead-time distribution.;It should be noted that this dissertation was initially simulation based. A large computer program with over five hundred lines of code was first developed to generate the solution to the problem through a search procedure and, then, to compare it with a simulated inventory model that relaxed the assumption of non-interchangeability. The numerical results obtained from printouts gave us insight to many of the problems addressed in this dissertation and eventually led to the analytical work that is described here.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
-
Program
-
Business
