Date of Award

3-14-2014

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Operational Sciences

First Advisor

Jeffery D. Weir, PhD

Abstract

Analysts at USTRANSCOM are tasked with providing vehicle mixtures that will support the distribution of requirements as provided in the form of TPFDD. An integer programming model exists to search for optimal solutions to these problems, but it is fairly time consuming, and produces only one of potentially several good quality solutions. This research constructs a number of heuristic approaches to solving the TDP. Two distinct shipping methods are examined and applied through both constructive and probabilistic vehicle assignment processes. Multistart metaheuristic approaches are designed and used in conjunction with the constructive and probabilistic approaches. Random TPFDDs of size 20, 100 and 1000 are tested, and solutions are compared to those obtained by the integer programming approach. The heuristic models implemented in this research develop feasible solutions to the notional TPFDDs in less time than the integer program. They can very quickly identify a number of good quality solutions to the same problem.

AFIT Designator

AFIT-ENS-14-M-25

DTIC Accession Number

ADA598850

Download

Share

COinS