DYNAMIC MODELING OF TRAFFIC NETWORKS IN ROUTE GUIDANCE FOR SYSTEM TIME OPTIMAL AND USER TIME EQUILIBRIUM

The paper presents and demonstrates an alternative approach to dealing with queues in the dynamic traffic assignment. This approach is mesoscopic in nature, as it combines the features of both macroscopic and microscopic dynamic traffic assignment/simulation methods. The new approach utilizes a simulation-oriented logic to trace the growth and decay of oversaturation queues by using vehicle packets. In addition, it incorporates an optimization procedure, which uses a successive average logic to achieve the actual time anticipatory user equilibrium and a tedious search method to find the approximate system time solution. Within the proposed approach, each vehicle packet, which can be any pre-specified size, is assigned into a minimum path that is associated with the determined optimal multiple path weights/splits, as well as traced along its minimum path on a link by link basis in terms of its arrival time slice to each link, the actual link travel time during the arrival time slice, the queue size it meets at each critical point and the necessary time to traverse each queue. The ultimate purpose to develop this mesoscopic approach is to seek a compromise solution between the existing macroscopic and the microscopic dynamic traffic assignment/simulation models. The research attempts to find a method that can deal with as many complex network and traffic dynamics as possible without losing the least optimization capability.

Language

  • English

Media Info

  • Features: Figures; References;
  • Pagination: p. 198-209

Subject/Index Terms

Filing Info

  • Accession Number: 00724735
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Aug 9 1996 12:00AM