Braess Paradox Scenario


Intersection Statistics

intersection_stat.gif
The figure above shows the route choice of the vehicles at the node from which the two route alternatives of the initial network depart. During the initial iteration all vehicles take route through links 1 and 2. As the iterations go on more and more vehicles switch to the alternate route through links 1 and 9.

Routing Statistics

routing_stat.gif
The figure above shows the number of vehicles in the study area by simulation time. After relaxation the number of vehicles in the modified network is generally higher than in the original network. This is because the longer route alternatives which could hold a larger number of vehicles are less frequently used.

Travel Time Statistics

traveltime_stat.gif

The figure above shows the core statement of the Braess paradox: after relaxation the average travel time of the vehicles in the modified ("improved") network is significantly worse than the in the original network. The average travel time increases from approximately 1100 to 1600 seconds. The reduced throughput of the modified network results in a increased simulation time until all vehicles have left the system.

Note that the original network is slightly above capacity which results in traffic jams in some of the iterations: iterations 99 and 100 have traffic jams whereas iterations 80 and 90 apparently do not have any.

Performance Statistics

performance_stat.gif
The figure above shows the real time ratio of PAMINA for this scenario.


since 05/01/2005 Aktualisiert am 01.05.2005 17:16 Uhr