Dynamic equilibrium assignment with microscopic traffic simulation

This paper presents a hybrid dynamic traffic assignment model by integrating the analytical dynamic user equilibrium technique and the micro-simulator paramics. The time-dependent path flows are generated by the analytical dynamic user equilibrium model, while the network attributes such as link and path travel times are simulated by the microsimulator. In particular, a path-based assignment method is developed in this paper to replace the off-the-shelf link-based assignment method in paramics. To assure the convergence of the proposed model, the method of successive averages is applied to update dynamic path flows from iteration to iteration. Numerical examples are also provided in this paper to test the performance of the proposed hybrid model and the solution algorithm.