Date of Award

3-22-2012

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Operational Sciences

First Advisor

Jeffrey K. Cochran, PhD.

Abstract

The objective of this study is to model operations at an airport passenger terminal to determine the optimal service capacities at each station given estimated passenger flow patterns and service rates. The central formulation is an open Jackson queueing network that can be applied to any USAF Air Mobility Command (AMC) terminal regardless of passenger type mix and flow data. A complete methodology for analyzing passenger flows and queue performance of a single flight is produced and then embedded in a framework to analyze the same for multiple departing flights. Queueing network analysis (QNA) is used because no special software license or methodological training is required, results are obtained in a spreadsheet model with computational response times that are instantaneous, and data requirements are substantially reduced compared with discrete-event simulation (DES). However, because of the assumptions of QNA, additional research contributions were required. First, arrivals of passengers are time-dependent, not steady-state. Theoretical results for time-dependent queue networks in the literature are limited, so a method for using DES to adjust for arrival time-dependency in QNA is developed. Second, beyond quality of service in the network, a key performance measure is the percentage of passengers who do not clear the system by a fixed time. To populate the QNA mean value system sojourn time, DES is used to develop a generic sojourn time probability distribution. All DES computations have been pre-calculated off-line in this thesis and complete a hybrid DES/QNA analytical model. The model is exercised and validated through analysis of the facility at Hickam AFB, which is currently undergoing redesign. For larger flights, adding a server at the high-utilization queues, namely the USDA inspection and security screening stations, halve system congestion and dramatically increase throughput.

AFIT Designator

AFIT-OR-MS-ENS-12-14

DTIC Accession Number

ADA561228

Download

Share

COinS