Dynamic traffic assignment with a node-based cell transmission model satisfying the link-level first-in-first-out principle

This repository contains the codes related to a research paper I've co-authored, whose abstract is reproduced below:

This paper develops node-based formulations for user equilibrium (UE) and system optimum (SO) dynamic traffic assignment (DTA) problems with departure time choice and route choices for general multiple origin-destination networks. Both the formulations are embedded with a new cell transmission model that satisfies the link-level First-In-First-Out (FIFO) principle. Because the formulations are node-based, the need for path enumeration is obviated, which results in considerable computational efficiency compared to the existing path-based models. While this advantage of node-based (or bush-based) models has been widely accepted in the literature of static traffic assignment, the formulations of dynamic traffic assignment models have mostly been path-based. The present work first describes a node-based cell transmission model that satisfies the link-level FIFO principle, which is fit within a DTA framework that facilitates efficient computation of UE and SO solutions. Further contributions of the work include the introduction of a backpropagation algorithm to efficiently compute marginal costs and complementarity formulations of the problems. Finally, numerical results are presented to demonstrate the performance of the proposed models using two standard test networks, along with a discussion of their convergence.

For more details, please refer to the following:

Mayakuntla, S. K., & Verma, A. (2019, June 12). Dynamic traffic assignment with a node-based cell transmission model satisfying the link-level first-in-first-out principle. https://doi.org/10.31224/osf.io/85btc