AN INTEGRATED AGBM-DTA MODEL FOR OPTIMIZING THE TRANSPORTATION SYSTEM BENEFITS OF PERSONALIZED MONETARY AND NON-MONETARY INCENTIVES

Abstract

The employment of different types of incentives in transportation systems to form advanced transportation congestion management solutions has garnered significant attention recently. This dissertation develops an integrated and personalized incentive scheme to incentivize more system-beneficial travel and mobility options considering both monetary and non-monetary incentives.

In real-world case, when offered different travel options, the users usually choose the options with higher individual benefits, while the incentive providers aim to maximize system benefits. Therefore, conflicts occur between the agents (the users) and the principals (the incentive providers/system benefit optimizers) because the agents act solely based on their own interests. Thus, the principals provide both monetary and non-monetary incentives to minimize the agents’ efforts of altering their travel behaviors. To optimize system benefits, we continue investigating the allocation of monetary and non-monetary incentives in different scenarios with different incentive budgets.

Furthermore, to analyze and visualize the impact of different incentive policies, we propose to build an integrated AgBM-flashDTA model, namely an agent-based behavior model (AgBM) integrated with a dynamic traffic assignment model (DTA). The flashDTA is a newly developed DTA model with a tree-based framework to do traffic assignment. This novel assignment method can converge in seconds, much faster than other simulation tools, making the model a powerful tool for supporting real-time decision-making.

Finally, through a demonstrative case study for a large-scale transportation system in the Washington D.C. and Baltimore regions, the capability of the proposed scheme is highlighted with significant system-level savings, reasonable insights on individual travel behavior responses, as well as superior computation efficiency.