Route Guidance

A variety of approaches for route guidance, covering the entire spectrum from analytical to hybrid analytical-simulation models, have been proposed in the literature. Various assumptions are being made regarding the nature of information used for guidance generation, drivers' behavioral response to information and route guidance (including drivers' compliance to guidance).

With respect to the nature of information, some systems use predicted traffic conditions while others are based on current information. Most systems utilize network-wide information and a few utilize information on local traffic conditions. Assumptions on driver behavior vary widely. A common assumption, an extension of the static user equilibrium principle, is that users follow the shortest path to their destination. Some systems also assume that several classes of users exist and attempt to incorporate the effect of different behaviors (see [MIT(1996)] task B report for a review).

While researchers seem to be in overall agreement on the importance of basing guidance generation on predicted traffic conditions and incorporating realistic driver behavior, the performance, data and computational requirements of various approaches remain less studied. For example:

• How much improvement can be achieved if a system uses predicted rather than current information and what is the sensitivity to the quality of prediction?
• How different is the performance if a system generates guidance based on local rather than system-wide traffic conditions?
• How does the accuracy of input data (e.g., OD table and surveillance data) affect the quality of the guidance generated and the effectiveness of the ATIS?
• How robust is the system with respect to user behavior?

• Reactive Route Guidance
• Predictive Route Guidance
  • Rolling Horizon Model
    • Rolling horizon length
    • Rolling horizon step size
    • Guidance Resolution
    • Computational delay
  • Network State Estimation
  • Network State Prediction
    • Dynamic OD prediction:
    • Traffic prediction:
  • Guidance generation