Scenarios for the Traffic Microsimulation Suite PAMINA
Purpose
A scenario for PAMINA is a tested setup of files usually comprising the following components:
- a set of network files
- an initial routefile
- a makefile
- a Pamina configuration
- Gnuplot evaluation files
A scenario comes as a gzipped tar archive and must be installed into the scenarios subdirectory of the Pamina main directory.
Available Scenarios
- Braess Paradox:
This scenario contains a simple network in two different versions. The "original" network offers two equivalent alternative routes for going from A to B. On each alternative the vehicles have to go through a short stretch of road with a small capacity and a longer road with a large capacity. After replanning the stream of vehicles is distributed more or less equally onto the two alternatives. After the system has reached equilibrium the network is modified to include a third alternative. A vehicle taking that route has to travel a much shorter distance reducing its travel time noticeably. Unfortunatly, it has to go through BOTH stretches of low capacity of the first two alternatives. That way, the load on these two streches are pushed above capacity and after replanning the overall performance measured as the mean travel time of all vehicles is actually reduced. For a more detailed description of this paradox please see here.
See statistics plots generated with PAMINA.
- Dallas / Fort Worth Study Area
In this scenario the route planning is done for the complete network of the Dallas/Fort Worth metropolitan area based on a initial route set used in the TRANSIMS case study. Permission to use the street maps and the initial route plan was granted by Mr. Ken Cervenka of the North Central Texas Council of Governments on April 14th, 2005.
The routes of this route set are "trimmed" to the extent of the actual simulation area comprising a rectangle of approximately 11 by 13 kilometers and 2276 lane kilometers (resulting in 303500 CA sites).
During the initial iteration we do complete planning for all routes in the set based upon the free speeds of the street links. Then, we do 49 iterations with a replannning fraction of 5%, each time feeding back the actual travel times of the street links of the previous iteration. We see a drastical improvement of the throughput of the network as the iterations go on.
See statistics plots generated with PAMINA.
- Helsinki Metropolitan Area
In preparation.
Simulation Speed
TODO!
Downloads
| Name | Available Since Version | Latest Update | File Size | Link |
|---|---|---|---|---|
| Braess Paradoxon | 0.9 | 04/30/2005 | 6.254 | PaminaScenarioBraessParadox.tgz |
| Dallas / Fort Worth Study Area | 0.9.1 | 04/30/2005 | 77.224.415 | available upon request |
 |