Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/35186
Title: Modelling activity and travel choice behaviour : a network equilibrium approach
Authors: Fu, Xiao
Advisors: Lam, H. K. William (CEE)
Sumalee, Agachai (CEE)
Keywords: Choice of transportation
Urban transportation
Issue Date: 2015
Publisher: The Hong Kong Polytechnic University
Abstract: Travel demands are derived from the desire of individuals to participate in various activities such as home, work, shopping, etc. Individuals' travel choice and activity choice, together with the interdependence of activity and travel scheduling, should be comprehensively investigated by means of transport modelling. Network equilibrium models with taking account of congestion effects, which provide valuable insights into understanding individuals' activity and travel choice behaviour, are widely used for long-term transport planning.In densely populated urban areas such as Hong Kong, multi-modal trips have been increasing in magnitude in recent years. This situation is similar in many fast- growing cities in Asia. Hence finding equilibrium results in congested multi-modal transport/transit networks is an important issue in long-term transport planning. The research presented in this thesis is novel in its aim to address the activity and travel choice behaviour simultaneously with consideration of congestion effects such as the crowding at activity locations and within transit vehicles in multi-modal transport/transit networks.In the literature, most existing network equilibrium models for travel behaviour analysis adopt a trip-based approach. In the trip-based approach, trips are adopted as the basic unit of analysis, and trip chains made by individuals are considered as separable and independent entities.The first contribution of this thesis is that a trip-based network equilibrium model is proposed to model travel choice behaviour in congested multi-modal transport networks under demand uncertainty. In the proposed trip-based model, crowding discomfort in transit vehicles, boarding congestion effect, and congestion impact of road traffic are explicitly modelled. The stochastic bus frequency derived from unstable road travel time is particularly investigated. The impacts of demand uncertainty on passenger flows and travel times are effectively captured. Individuals' route and mode choice behaviour under travel time uncertainty are intensively explored.The trip-based network equilibrium model, however, ignores the underlying motivation of trip making, and cannot reflect the linkages between activities and travels. To understand the limitation and shortcoming of the trip-based approach, the above trip-based proposed model is extended to an activity-based approach. The activity-based approach enables an integrated investigation into the activity-travel scheduling mechanism, i.e. what activities to be conducted, in what sequence, when and for how long, when each trip starts, which transport mode/route is to be used, and how the activities and travels interrelate in congested multi-modal transport/transit networks. As previous studies have shown that crowding discomfort has an important effect on individuals' choice of transit service for long-term planning, the in-vehicle crowding discomfort is considered in the proposed activity-based model particularly for congested transit networks in Asia.The second contribution of this thesis is the proposal of an activity-based network equilibrium model to solve the daily activity-travel pattern (DATP) scheduling problem. The resultant DATP choice reflects individuals' various activity choices (e.g.activity sequence, start time and duration), travel choices (e.g. departure time, transfer,and route/mode), and the relationship between activity and travel choice behaviour in congested multi-modal transit networks.
As a pioneering endeavour, the proposed activity-based network equilibrium model extends existing theories by developing an integrated framework which incorporates the flexible activity sequence and duration, the stochastic effects of activity utility, together with the route/mode choice under network congestion. A novel activity-time-space multi-modal super-network platform is constructed to explicitly address the relationship between activity choices and travel choices in time and space coordinates in a congested multi-modal transit network. By using the proposed super-network platform, the time-dependent DATP scheduling problem can be converted into a static traffic assignment problem.A number of empirical studies have investigated the recurrent effects of adverse weather on individuals' DATP choice and such effects are obviously greater in cities which suffer frequent rainy periods. The long-term transit planning for areas with high average annual rainfall should be considerably different from the planning for areas with less rainfall. Thus, clearly, particularly in areas such as Hong Kong, rain effects should be considered when modelling individuals' activity and travel choices. An activity-based network equilibrium model for scheduling DATPs in multi-modal transit networks under adverse weather conditions (with different rainfall intensities) is proposed, which is the third contribution of this thesis. The interdependency of individuals' activity/travel choices and weather conditions are intensively investigated in congested multi-modal transit networks. As vehicle capacity and frequency of different transit modes are influenced by adverse weather conditions, in-vehicle crowding discomfort taking account of adverse weather impacts is specifically considered in the proposed model. The effects of adverse weather on different transit modes and different activities are also explicitly investigated.It should be noted in the above three network equilibrium models that the activity and travel choice behaviour of individual is assumed to be independent, so called one-individual level. As travel surveys indicate, joint participation in activities and travels represent a substantial portion of individuals' DATPs. Most existing studies on activity-based network equilibrium models, however, are confined to the one-individual level. Less attention has been given to the interdependence between individuals' joint activities and travels. Obviously, there is a need to investigate the effects of the joint activity-travel pattern (JATP) choice for long-term transit planning.The fourth contribution of this thesis is the development of an activity-based network equilibrium model which can solve two-individual JATP scheduling problem in congested multi-modal transit networks. The proposed JATP scheduling model extends existing theories by developing a unified framework to capture both independent and joint activity/travel choices in congested multi-modal transit networks. To capture the effect of activity location capacity in long-term transit planning, the proposed model is extended to incorporate the crowding discomfort at activity location. A measure of JATP utility is proposed to incorporate the joint travel benefit. Individuals' preference towards joint travel is explicitly examined by the proposed model, and the impacts of joint travel benefit on individuals' independent and joint activity/travel choices are intensively explored.Network equilibrium models are capable of predicting traffic flow patterns subject to network congestion phenomena. In this thesis, four network equilibrium models are proposed to comprehensively investigate individuals' activity and travel choice behaviour in congested multi-modal transport/transit networks. In the proposed models, different congestion effects are considered such as in-vehicle crowding discomfort, road traffic congestion, and crowding at activity location. The proposed models offer the flexibility and feasibility to comprehensively consider different congestion effects in multi-modal transport/transit networks for future extensions. The ultimate aim of the proposed network equilibrium models is to make valuable contributions to the new avenue of research on activity and travel choice behaviour for design of multi-modal transport networks and evaluation of alternative transport systems with consideration of their congestion effects.
Description: PolyU Library Call No.: [THS] LG51 .H577P CEE 2015 Fu
xxvi, 176 pages in various paging :illustrations ;30 cm
URI: http://hdl.handle.net/10397/35186
Rights: All rights reserved.
Appears in Collections:Thesis

Files in This Item:
File Description SizeFormat 
b28163333_link.htmFor PolyU Users203 BHTMLView/Open
b28163333_ir.pdfFor All Users (Non-printable)2.24 MBAdobe PDFView/Open
Show full item record

Page view(s)

57
Last Week
3
Last month
Checked on Mar 26, 2017

Download(s)

20
Checked on Mar 26, 2017

Google ScholarTM

Check



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.