A methodology to design multimodal public transit networks : procedures and applications

This thesis deals with public transit network design problem, one of the core issues facing the existing planning and operation of public transit systems. More specifically, this problem involves designing an efficient/optimal set of routes and their associated frequencies for public transit network...

Full description

Saved in:
Bibliographic Details
Main Author: Ul Abedin, Zain
Other Authors: Wang Zhiwei, David
Format: Thesis-Doctor of Philosophy
Language:English
Published: Nanyang Technological University 2019
Subjects:
Online Access:https://hdl.handle.net/10356/136540
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:This thesis deals with public transit network design problem, one of the core issues facing the existing planning and operation of public transit systems. More specifically, this problem involves designing an efficient/optimal set of routes and their associated frequencies for public transit networks, given certain constraints. A novel route overlap-based approach is developed for this design undertake considering the perspectives of passengers and agencies. The concept of route overlap entails the ratio of common segments/arcs that each route shares with the remaining set of the routes of the transit network. Varying this ratio results with a variety of transit-network configurations such as a sparse network with lower ratios and dense network with higher ratios. The main part of the developed methodology consists of four major components. It starts with transit route creation and network construction where, at first, feasible routes are created by using the traditional k-shortest path algorithm, and then route overlap concept is used to design a public transit network. The second component is the analysis of assigning transit demand including headway derivation; this is used to assign demand onto the transit network and to generate travel times and suitable transit modes. The third component is the holistic evaluation of the transit network; this is done by using performance metrics to reflect the viewpoints of passengers, agencies and authorities. The last component is a metaheuristic search engine employed to explore feasible search space for attaining improved transit network configurations from both passengers and agencies perspectives. The new methodology undergone testing of different networks including benchmark and real-life networks. The outcomes of the experiments are used for the evaluation and validation of the proposed methodology including a comparison with other research studies. The results show that instead of using many overlapping redundant routes, it is more efficient and productive to have lesser number of routes with a reduced number of overlapping route segments; this will be more suitable and efficient for both passengers and agencies. Finally, the viability of the developed methodology is examined for the Singapore’s public transit network; the results suggest that this network can be optimised further with even lesser number of routes.