Flexible transit network design with and without branching under spatially heterogeneous demand

While public transportation systems are usually designed with fixed routes, this work presents an alternative flexible-route transit system. Flexible transit vehicles do not operate on fixed paths but travel within predetermined areas in response to trip demand in order to provide door-to-door service. The main advantage of this system is that passenger access time to and from transit stops is removed. To design the optimal route layout and service operation, continuum approximation is used to reduce the computation burden and formulate the problem in terms of a few decision variables. Unlike many continuous models, passenger distribution is not assumed to be homogeneous over space. Since travel patterns are typically not uniform in urban and suburban areas, this thesis will consider a heterogeneous passenger distribution. In order to adapt to both global and local demand variations, several transit system designs will be investigated. Thus, the purpose of this thesis is (i) to investigate the benefits of Daganzo (2010a) hybrid structure over grid structure under heterogeneous demand with flexible routes; and (ii) to investigate the benefits of allowing branching local tubes within the transit system. For (i), we derive the agency and user cost metrics of the proposed models and seek optimal network layout, service area of each bus and bus headway, to minimize the total generalized cost. For (ii), we use the framework provided by Ouyang et al. (2014) and the power-of-two concept from Roundy (1985) to design a grid flexible transit network with local tubes. The same cost metrics as in (i) are derived on a local scale. Considering a low-to-moderate demand level and several spatially heterogeneous demand distributions, it is found out that hybrid structure is beneficial over grid structure, and that transit network with local tubes allows a reduction of the system cost, with respect to a homogeneous transit network. A sensitivity analysis is performed on the branching structure design. It is found that branching does not depend on the total number of passengers. Finally several future research leads are presented to enhance the transit network design.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Civil Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2015

Author dc:creator

• Petit, Antoine

Contributors dc:contributor

• Ouyang, Yanfeng

• flexible transit
• hybrid network
• branching

Statement dc:rights

• Copyright 2014 Antoine Petit

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/73106

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/73106