Proposal Summary and Objectives

Freight traffic congestion and delays have been a pressing issue nationally especially in the wake of the pandemic when the supply chain is stressed. Severe congestion and delays have been observed in and around major ports and corridors. Freight traffic plays an important role in supporting efficient logistics operation and national productivity. Therefore, there is a significance in incorporating freight traffic in the general traffic operations and control. This proposal aims at developing new network signal algorithms when freight traffic is considered. The goal is to have algorithms that can serve both the passenger and freight travelers. In general, intersection signal control has seen immense progress in the past decades. However, optimal models/theories and algorithms are still lacking, as compared with the vast development and application of information technology. The PI, several years ago has published a methodological paper on exploring the optimal structure of intersection control in Transportation Research Part B with an algorithm named DORAS. That algorithm was tested for isolated intersections only, although it carries the potential to extend the model into the case of general network. This proposal plans to build on the results therein and further explore the optimal framework for the case of network. This exploration will naturally piggyback with another parallel research effort that a PhD student of the PI’s has been conducting as his dissertation, which mainly explores a theoretical framework of Lyapunov’s dynamic control theory coupled with max pressure method with reinforced learning. We hope to complete a full set of methodologies with the addition of the newly proposed algorithm to explore and hope to compare them together at the end of the proposed project.

Funding Amount:
Status: Active
Duration: Sep 1, 2021 - Jul 30, 2022