ANALYTICAL DYNAMIC TRAFFIC ASSIGNMENT MODELS: FEATURES AND SOLUTION EXPERIENCE

The dynamic traffic assignment (DTA) problem is studied both through analytical and simulation-based approaches. This paper proposes an analytical DTA model intended for off-line Advanced Traffic Management and Information Systems (ATMIS) evaluation and application. A link-based variational inequality (VI) model for the dynamic user-optimal (DUA) traffic assignment problem and a diagonalization algorithm are proposed and implemented on a realistic, large-scale traffic network. The proposed solution procedure can approximate route choices based on anticipatory and non-anticipatory traffic conditions. Adjustments of link capacities (caused externally and internally) are specifically considered in the solution algorithm in order to capture queue formation and dissipation. Computational results and concluding remarks are presented. A major contribution of this research is that the proposed model, solution procedure, and associated features demonstrate capabilities for realistic network-wide applications with ATMIS implications.