A FORMULATION AND SOLUTION ALGORITHM FOR A MULTI-CLASS DYNAMIC TRAFFIC ASSIGNMENT PROBLEM

Dynamic traffic assignment (DTA) has been a topic of substantial research during the past decade. While DTA is gradually maturing, many aspects of DTA still need improvement, especially regarding its formulation and solution capabilities. In order to model the impact of Advanced Transportation Management and Information Systems (ATMIS), especially route guidance and other information provision systems, it is necessary to develop a set of multi-class traffic models to acknowledge the fact that there are different classes of users of the transportation system, and that they respond differently to traffic information. At least, the model should be able to differentiate travelers or vehicles who receive real-time traffic information versus those who don't. This paper aims to advance the state-of-the-art in DTA modeling. In this paper, an analytical approach is developed to model multiple classes of users of the transportation system. It is a step forward of the long-term efforts in analytical DTA modeling, which distinguishes itself from simulation-based DTA models. Specifically, the users are divided into three classes: (i) fixed route travelers; (ii) stochastic dynamic user-optimal (SDUO); (iii) dynamic user-optimal (DUO). The property of each class is defined and integrated into one modeling framework through a Variational Inequality (VI) approach. Subsequently, a solution algorithm for the formulation is discussed. This algorithm uses a combination of various solution techniques, such as relaxation, Frank-Wolfe and Method of Successive Averages (MSA). It is applied to four scenarios to verify the correctiveness of the solutions obtained. (A) For the covering abstract see IRRD 886400.