Book/Dissertation / PhD Thesis

FZJ-2022-01956

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Operational Navigation of Agents and Self-organization Phenomena in Velocity-based Models for Pedestrian Dynamics

Xu, Q. (Corresponding author)FZJ*

2022
Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag Jülich
ISBN: 978-3-95806-620-5

Please use a persistent id in citations: http://hdl.handle.net/2128/31123

Abstract: While moving in space, pedestrians often adjust their direction of movement and/or their speed to avoid collisions with others and obstacles. This steering process is influenced by physical factors from the environment, as well as psychological factors of pedestrians such as motivation. Therefore, when modeling the movement of pedestrians especially for reproducing self-organization phenomena, it is important to consider these factors. This cumulative dissertation includes four publications related to velocity-based models for pedestrian dynamics. Three of them study the navigation of pedestrians and related self-organization phenomena, and the remaining one is an applied study related to the control measures adopted by German supermarkets during the COVID-19 pandemic. Velocity-based models consider pedestrians as particles also called agents and describe their movement at the operational level, by means of first-order differential equations. Velocity-based models, contrary to cellular automata, are continuous in space. Moreover, the new position of pedestrians is determined directly by a velocity function instead of an integrating of acceleration in force-based models. In publication I, a velocity-based model that considers several basic behaviors of pedestrians is proposed and validated with the fundamental diagram of unidirectional pedestrian flow. Besides, the effect of agents’ shape on the overall dynamics is studied. Although this basic model is able to guarantee the volume exclusion and reproduce the fundamental diagram of unidirectional pedestrian flow, it does not perform well incomplex scenarios, where self-organization phenomena occur. Therefore, in publication II and III, the previously developed basic model is used to quantitatively study clogging in bottleneck scenarios and lane-formation in bidirectional flow scenarios, respectively. In addition, in publication III, an anticipation mechanism is introduced into the basic model to describe lane-formation in bidirectional flow scenarios more realistically.

Note: Dissertation, Univ. Wuppertal, 2022

---

Contributing Institute(s):

1. Zivile Sicherheitsforschung (IAS-7)

Research Program(s):

1. 5111 - Domain-Specific Simulation & Data Life Cycle Labs (SDLs) and Research Groups (POF4-511) (POF4-511)

Appears in the scientific report 2022

---

Database coverage:
;

---