Year
2024
Degree Name
Doctor of Philosophy
Department
School of Computing and Information Technology
Abstract
Walking and cycling are essential and important components in facilitating the sustainability of multimodal transport systems and forming a healthy lifestyle for the communities. Safety concerns of vulnerable road users bring great challenges to road infrastructure design and policy regulations development. To better understand vulnerable road users’ behaviours and protect their safety, a growing amount of research has been conducted on crash analysis, crowd dynamics and behaviour modelling over the past decades. However, how to further reduce their safety risks and improve their comfort levels under various circumstances still needs to be investigated. Especially, the increasing prevalence of emerging mobilities (e.g., e-bicycle and e-scooter) and the long-term impact of COVID-19 pandemic potentially posed great threats on vulnerable road users’ safety and comfort levels. To fill the research gap, this thesis aims to conduct safety analysis and behaviour modelling of vulnerable road users under complicated and mixed road environments. Unobtrusive observation methods were mainly adopted to collect multi-source dataset in real-life scenarios to extract road users’ trajectories, interactive behaviours and surrounding environments. The investigated scenarios included unidirectional pedestrian flow with focus on overtaking behaviour and COVID-19’s impact, crossing flow of pedestrians and cyclists at different types of bus stops with interactions and conflicts, and bidirectional mixed flow of pedestrians, cyclists and e-cyclists at non-segregated and segregated signalised intersections. In addition, pedestrian and cyclist crashes at micro- and macro-levels in the Australian Capital Territory (ACT) in Australia were explored. In total, more than 20,000 vulnerable road users (including pedestrians, cyclists and e-cyclists) were observed and analysed in this thesis. Statistical analysis was utilised to reveal the behaviour differences among heterogenous vulnerable road users, such as road user characteristics, traffic features and road environments. Microscopic models were developed to describe and reproduce road users’ movement patterns (e.g., speed, route choice, social distancing, and collision avoidance); and statistical models were adopted to estimate the effects of multiple influencing factors on safety risks of road users and operation efficiency of transport facilities (e.g., conflict occurrence, conflict severity, lag-time and evasive actions). The research findings are expected to provide valuable information to reveal reasons behind traffic crashes, and predict potential safety risks and congestion issues of vulnerable road users. Based on the research outcomes, multiple recommendations on transport infrastructure design and policy making on traffic regulations are proposed. This research is able to provide an opportunity to reduce the safety and comfort issues of vulnerable road users, improve the effectiveness of policy regulations and planning strategies, and create a safe and vulnerable road user-friendly road environment.
Recommended Citation
Zhang, Cheng, Safety Analysis and Behaviour Modelling of Vulnerable Road Users, Doctor of Philosophy thesis, School of Computing and Information Technology, University of Wollongong, 2024. https://ro.uow.edu.au/theses1/1785
FoR codes (2008)
090507 Transport Engineering, 120506 Transport Planning, 120508 Urban Design
Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.