摘要: |
Route (or path) planning is a core optimization problem to address for efficient and intelligent transportation in various transportation systems. While the algorithms designed for efficient and accurate route planning in transportation networks is extensive, to the best of the research team's knowledge all existing solutions focus on planning optimal routes for individual travelers. With this approach, "optimality" is defined based on a criterion that captures best interest(s) of individual travelers rather than those of the transportation network as a whole. Although popular, this definition of optimality is not necessarily aligned with the strategic goals of the USDOT, which demand optimal utilization of the transportation network in terms of performance measures such as overall mobility and environmental sustainability.
With the previous Mountain-Plains Consortium (MPC) project, the research team addressed this misalignment by introducing system-optimal route planning, an alternative approach to route planning where optimality of the routes is defined based on their impact on overall utilization of the transportation network rather than benefits of individual users. In particular, these solutions leverage two big data methodologies, namely, guaranteed approximation and distributed and parallel computation, to scale up route planning for practical applications.
With this proposal, the team plans to extend their system-optimal route planning solutions to consider scenarios where certain user constrains ought to be enforced for valid system-optimal route planning. In particular, the team intends to develop constrained system-optimal route planning solutions for fleet route planning, ridesourcing and ridesharing, and perform an extensive simulation-based comparative analysis to evaluate performance of the proposed solutions versus existing state-of-the-art solutions. |