Please use this identifier to cite or link to this item:
Title: Modeling activity and travel choice behaviour for land use and transport optimization problems
Authors: Ouyang, Lianqun
Keywords: Transportation -- Mathematical models.
Choice of transportation -- Mathematical models.
Trip generation -- Mathematical models.
Hong Kong Polytechnic University -- Dissertations
Issue Date: 2014
Publisher: The Hong Kong Polytechnic University
Abstract: There has been a growing recognition in recent years that the transport models which use the activity-based approach could provide better understanding of user travel choice behaviour, when compared to the conventional transport models which use the traditional trip-based approach. The traditional trip-based transport models usually focus on a single trip with the assumption that all trips are independent to each others. Activity-based transport models, however, investigate the underlying activity choices which generate trips, interdependence of trips, scheduling of activities and trips regarding time and space, as well as relationships between inter-personal activities and trips. Thus, as can be seen, activity-based transport models describe user travel choice behaviour comprehensively. In addition, due to the strong interactions between land use and transport, it is widely believed that there is a need to develop integrated land use and transport models to simultaneously model user (residential/employment) location and travel choice behaviours. In view of the potential merits of the activity-based approach in modeling user travel choice behavior together with their close relationships between user residential/employment location and activity choices, it is highly desirable to move forward by developing integrated land use and transport models using the activity-based approach. Land use and transport have a strong interactive relationship; hence the development of sustainable land use and transport optimization models could be an important step to relieve traffic congestion problems in fast-growing cities. Examples of problems resulting from such growth are currently seen in China. The growing activity demands due to such rapid development, in addition, are causing grim traffic air-pollution problems. Urban development resulting from too rapid economic growth and without the foresight of comprehensive and well structured planning in efficient clean transport can negatively affect human physical and mental health, and unnecessarily, interrupt economic progress. How to balance the economic development and environmental pollution associated with sustainable land use and transport policies, is frequently one of the most acute problems with which authorities of fast-growing cities are concerned. With the above urban problems in mind, the overall focus of this research is developing models for modeling user travel choice behaviour more accurately so that transport-related or land use polices which enable efficient and clean movement within a city could be explored. In this regard, this research offers three major contributions. 1. The user daily activity scheduling problem under two scenarios (in congested road networks and in general road networks) is studied; two activity-based transport models are proposed and provide the first contribution of this research. The first model focuses on the user daily activity scheduling network equilibrium problem in congested road networks. It is formulated as an equivalent user equilibrium (UE) assignment problem for an expanded Activity-Time-Space (ATS) queuing network. The second model is an extension of the first and deals with the user daily activity scheduling network equilibrium problem in general road networks (i.e. both congested and non-congested road networks) and incorporates users' different perception variations on activity utilities and travel times. It is shown to be a stochastic user equilibrium (SUE) assignment solution to the urban road network for solving the user daily activity scheduling network equilibrium problem.
In the proposed two activity-based transport models, the motivation of making trips, unlike the previous trip-based transport models, the relationships between trips, and the time and space coordinates are investigated explicitly. In addition, the activity sequences and activity durations are endogenously generated in the proposed models. Thus, the user travel choices are examined in a more comprehensive framework by comparing to the relevant activity-based transport studies appearing in recent years in the literature. Some properties of the equilibrium solution are also investigated and discussed. Efficient solution algorithms without the prior Daily Activity-Travel Pattern (DATP, i.e. user daily activity scheduling result) enumeration are developed for solving the above two models. In addition, the proposed activity-based transport model for general road networks allows different perception error variations of utilities/disutility across various activities and travel. It is noted that these variations are usually assumed to be identical in the previous activity-based related studies. 2. The development of an activity-based land use and transport model to investigate the combined user residential location and travel choice problem is the second contribution of this research. The users' combined residential location choice and travel choice problem is formulated as a stochastic user equilibrium network model. Users are assumed to pursue the perceived maximum difference between benefit and cost of selecting a residential location. The benefit of selecting a residential location is equal to the total utility obtained from participating in activities minus the travel costs needed for activity participations in the study time horizon. The cost of choosing a residential location is the housing price or rent in the study time horizon. In this research, the developed activity-based land use transport model extends the existing relevant models in the literature by considering both user residential location choices and user travel choices with the use of the activity-based approach. The close interrelationship between user residential location choice and user DATP choices is revealed in this developed model. The inconsistency problem that may appear in some existing relevant models is avoided in the developed model. This extension offers a new avenue of research on network equilibrium analysis for examining the interactions between land use and transport. 3. As far as the third contribution of this research is concerned, a novel activity-based land use and transport optimization model using the bi-level programming technique is proposed to solve the sustainable land use and transport optimization problem. The upper level of the proposed model is to optimize the residential/employment location developments and road link capacity expansions with a budget constraint so that the maximum network social welfare is achieved while satisfying the network-wide traffic emission control level. The lower level of the proposed model is a network equilibrium model which models combined user residential/employment location and daily activity/travel choice (i.e. DATP choice) behaviour. The user location and travel choice behaviours are accurately captured with the use of the activity-based approach in the lower level of the proposed model. The proposed bi-level programming model is able to optimize the land use development and transport system improvement in a consistent way.
Description: xxxiii, 190 pages : illustrations ; 30 cm
PolyU Library Call No.: [THS] LG51 .H577P CEE 2014 Ouyang
Rights: All rights reserved.
Appears in Collections:Thesis

Files in This Item:
File Description SizeFormat 
b27575275_link.htmFor PolyU Users 203 BHTMLView/Open
b27575275_ir.pdfFor All Users (Non-printable) 2.07 MBAdobe PDFView/Open
Show full item record

Page view(s)

Last Week
Last month
Checked on Jan 15, 2017


Checked on Jan 15, 2017

Google ScholarTM


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