Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [sumo-user] Realistic Simulation with traffic data and calibration

I recommend working with the bigger network (tuning individual junctions as necessary). It is always easier to cut out a smaller scenario from a bigger one than assembling a big scenario from smaller ones.
The effect of multiple intersections is noticeable when some junctions are over-saturated. Then the jams grow upstream and affect other junctions as well.
However, if you don't have high traffic and the junctions are far away from each other than it doesn't matter too much.
For pedestrians, it depends:
- pedestrians may block turning traffic that has green when the pedestrians can cross.
- pedestrian phases (i.e. a pedestrian scramble) may reduce the green time for cars

Am Fr., 17. Apr. 2020 um 16:09 Uhr schrieb Tetris <schmelter@xxxxxxxxxxxxxxxxxxx>:
I tried the systematic approach for one of the junctions and it was totally
fine, no traffic jam at all. So I will take a closer look on all the other
junctions one by one.

In general, is it better to have a network with only junctions in it that
have turncounts or is it better to have a bigger network to create realistic
I am asking because I am going to change every connection and tlsIndex in
Netedit and if I use the minimal network I would have to do it all over
again for the bigger network if the bigger network leads to a more realistic
traffic flow. I guess the best approach is to change every junction in
Netedit for the bigger network and choose afterwards.

Another question: is it necessary to create pedestrian lights and phases for
them to get realistic traffic flows or is it sufficient to define traffic
lights just for vehicles? There are no pedestrians in the simulation and I
am just interested in results for travel time and acceleration of my egoCar.

Sent from:
sumo-user mailing list
To unsubscribe from this list, visit

Back to the top