Simulation and optimisation of dynamic multimodal traffic network with shared mobility systems
File(s)
Author(s)
Jeamchantamook, Peeranut
Type
Thesis
Abstract
Rapid urbanisation continues to intensify traffic congestion, a challenge compounded by the “last-mile” problem and limited understanding of shared mobility systems. Existing research frequently employs isolated models, failing to capture complex interdependencies spanning network effects, matching algorithms, environmental outcomes, HOV lane performance, pricing, and public transport interactions.
This thesis addresses this gap by developing a comprehensive, computationally efficient, macroscopic multimodal traffic simulation framework. It integrates ridesharing, taxi-sharing, private, and public transport within a unified environment, incorporating within-day and day-to-day dynamics. A novel macroscopic matching algorithm, formulated using linear programming and hierarchical clustering, is introduced to efficiently accommodate large-scale demand while maintaining tractability.
Simulations demonstrate that integrating matching algorithms with route choice models is essential, as isolated analyses overestimate uptake by ignoring induced demand. Results show shared mobility systems can reduce congestion by up to 17.5% while improving network speeds by 4.14%. Substantial reductions in emission-related indicators like vehicle-kilometres travelled are also observed.
Policy design is decisive: maximising Vehicle Kilometres Travelled Saved (VKTS) reduces congestion by 17.22%, whereas profit-maximisation achieves 15.9%. Surge pricing raises revenue, but excessive factors may worsen congestion. HOV lanes promote ridesharing but require calibration to avoid displacing congestion. The framework also highlights the competitive relationship between shared mobility and public transport, noting bus ridership declines in low-congestion areas. Finally, the model determines optimal fleet sizes under different objectives.
These findings demonstrate that the proposed framework provides planners and researchers with an advanced tool for investigating shared mobility's long-term impacts. It supports the design of data-driven strategies to reduce congestion, promote sustainability, and integrate shared mobility effectively into urban systems.
This thesis addresses this gap by developing a comprehensive, computationally efficient, macroscopic multimodal traffic simulation framework. It integrates ridesharing, taxi-sharing, private, and public transport within a unified environment, incorporating within-day and day-to-day dynamics. A novel macroscopic matching algorithm, formulated using linear programming and hierarchical clustering, is introduced to efficiently accommodate large-scale demand while maintaining tractability.
Simulations demonstrate that integrating matching algorithms with route choice models is essential, as isolated analyses overestimate uptake by ignoring induced demand. Results show shared mobility systems can reduce congestion by up to 17.5% while improving network speeds by 4.14%. Substantial reductions in emission-related indicators like vehicle-kilometres travelled are also observed.
Policy design is decisive: maximising Vehicle Kilometres Travelled Saved (VKTS) reduces congestion by 17.22%, whereas profit-maximisation achieves 15.9%. Surge pricing raises revenue, but excessive factors may worsen congestion. HOV lanes promote ridesharing but require calibration to avoid displacing congestion. The framework also highlights the competitive relationship between shared mobility and public transport, noting bus ridership declines in low-congestion areas. Finally, the model determines optimal fleet sizes under different objectives.
These findings demonstrate that the proposed framework provides planners and researchers with an advanced tool for investigating shared mobility's long-term impacts. It supports the design of data-driven strategies to reduce congestion, promote sustainability, and integrate shared mobility effectively into urban systems.
Version
Open Access
Date Issued
2023-06-03
Date Awarded
2026-02-01
Copyright Statement
Attribution-NonCommercial 4.0 International Licence (CC BY-NC)
License URL
Advisor
Ochieng, Washington
Sponsor
Thailand
Publisher Department
Department of Civil and Environmental Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)