Date of Award

3-2003

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Operational Sciences

First Advisor

Richard F. Deckro, PhD

Second Advisor

Victor D. Wiley, PhD

Abstract

Current infrastructure network models of single functionality do not typically account for the interdependent nature of infrastructure networks. Infrastructure networks are generally modeled individually, as an isolated network or with minimal recognition of interactions. This research develops a methodology to model the individual infrastructure network types while explicitly modeling their interconnected effects. The result is a formulation built with two sets of variables (the original set to model infrastructure characteristics and an additional set representing cuts of interdependent elements). This formulation is decomposed by variable type using Benders Partitioning and solved to optimality using a Benders Partitioning algorithm. Current infrastructure network models of single functionality do not typically account for the interdependent nature of infrastructure networks, Infrastructure networks are generally modeled individually, as an isolated network or with minimal recognition of interactions, This research develops a methodology to model the individual infrastructure network types while explicitly modeling their interconnected effects, The result is a formulation built with two sets of variables (the original set to model infrastructure characteristics and an additional set representing cuts of interdependent elements) This formulation is decomposed by variable type using Benders' Partitioning and solved to optimality using a Benders' Partitioning algorithm.

AFIT Designator

AFIT-GOR-ENS-03-14

DTIC Accession Number

ADA420865

Download

Share

COinS