Date of Award

Spring 5-2022

Document Type

Dissertation

Degree Name

Doctor of Engineering (D Eng)

Department

Computational Modeling & Simulation Engineering

Program/Concentration

Modeling and Simulation

Committee Director

John A. Sokolowski

Committee Member

Andrew J. Collins

Committee Member

Roland Mielke

Abstract

As pedestrian evacuations of buildings, outdoor venues, and special events occur, dynamic interactions between pedestrians and vehicles during egress are possible. To model pedestrian and vehicle evacuations, simulation models have evolved to incorporate more realistic crowd characteristics and behaviors to provide improved results. Past studies using modeling and simulation, specifically agent-based modeling, have explored pedestrian behaviors such as decision-making, navigation within a virtual environment, group formations, intra-group interactions, inter-group dynamics, crowd behaviors such as queuing and herding, and pedestrianvehicle interactions. These studies have led to relevant insights helpful to improving the accuracy of evacuation times for normal and emergency egress for preparedness and management purposes. As evacuating crowds are composed of individual pedestrians and social or familial groups, this project contributes to the study of pedestrian evacuation by exploring the incorporation of a subgroup not often considered in this area. Vulnerable individuals, such as the physically disabled, elderly, and children, can change the decision-making dynamic of a group leader while evacuating to safety. Current agent-based simulation models explore the intra- and inter- action and the effects on evacuation times; however, the vulnerable group members' influence is neglected. This project presents enhancements to pedestrian evacuations with vehicle interaction using an agent-based simulation model that includes the presence of vulnerable group members and their impact on decision-making and evacuation times. This project explores how changing behaviors due to the presence of vulnerable group members can collectively cause delays and increase evacuation times. Utilizing verification and validation methods, the credibility and reliability of the simulation model and its results are increased. The results show that the group leaders' decision-making differs when leading a vulnerable group versus a non-vulnerable group. Also, evacuation times increase with increased percentages of vulnerable groups within an evacuating crowd. A simulation tool can be utilized by end-users to explore specific evacuation scenarios in preparation for upcoming events and glean insight into how evacuation times may vary with differing crowd population sizes and compositions. Including vulnerable pedestrians in simulation models for evacuations would improve output accuracy and ultimately improve event training and preparation for future evacuations.

DOI

10.25777/19v2-nd88

ISBN

9798819394328

ORCID

0000-0002-0743-3779

Share

COinS