Anticipatory Traffic Information for Navigation Systems Framework As Example of Transport Planning Policy
J Casas, TSS-Traffic Simulation Systems and University of Vic; A Torday, TSS-Traffic Simulation Systems,ES
This paper presents a framework for supporting navigation systems and providing alternatives routes to the users as a transport planning policy. This information is calculated using a dynamic traffic assignment and a mesoscopic model.
The technological evolution of the navigation systems, combined with a centralised systems for providing traffic information to those navigation systems could be considered as a new alternative of transport planning policies.
Technological considerations aside, automotive navigation systems use the road database as the main means of providing the user with information about the sequence of directions or waypoints on the map. The role of the road database here is to define a directed graph encoded with geographic coordinates; each link on the graph represents a segment of streets or roads with its associated costs, based on travel time, which can be used to calculate the shortest route to the user?s destination.
When providing the user with information, the concept of cost is the most relevant element. The type of cost under consideration affects which routes might be suggested to the user. The concepts of cost are: a) Travel time calculated in free flow conditions, b) Traffic data considering the historical profiles, and c) Real-time traffic data available to online subscribers. In all cases the algorithm for calculating the shortest route is common for all users and is embedded inside the navigation system.
This approach has the following potential problems, depending on the concept of cost: a) Travel time calculated in free-flowing conditions does not reflect the current traffic state of the network,
b) Traffic time calculated with historical profiles does not represent the current traffic state due to unexpected incidents,
c) Travel time based on real-time traffic data can capture the unexpected incidents, but the rate of subscription to this service may produce secondary effects, such as artificial congestion. Known as The Flip-Flop Effect, this type of congestion occurs when all users are rerouted to the same area of the network when avoiding the observed incident because all of the users have the same algorithm for calculating the new shortest route.
The framework proposed in this paper considers the real-time traffic state and the incidents present in the network. Using a centralised system, the proposal is to apply a dynamic traffic assignment based on a dynamic user-equilibrium and a mesoscopic simulation model for calculating the equilibrium situation. This would then allow a new route to be sent to each individual subscriber user, applying a random selection from the identification of all possible alternatives routes in a equilibrium situation, and maintain the equilibrium. The computational results using this framework evaluate the new system?s performance according to the service?s subscription rate.
Association for European Transport