Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [sumo-user] Arriving to a traffic break-down and observing dynamic shockwaves ( / phantom traffic jams)

Thank you very much for your answer Jakob, you helped me a lot!

I have 3 more questions:

1.       I have noticed that the attached test case which consists of a 1 hour 3000 veh/hr flow along 20 km of road performs just a single shock wave. Is it considered to be a realistic scenario?  If not - what can be done to make it more realistic? Which Vtype (or other) attributes do you recommend to edit in case of a 2 lanes scenario?

2.       I used the SUMO plot_trajectories tool and noticed that the graph looks very clean and maybe less realistic. How can I improve it?

3.       What is the reason that the road is divided into 100 sections? Can it affect the simulation results?


Sincerely, Dvir

‫בתאריך יום א׳, 3 באפר׳ 2022 ב-1:52 מאת ‪Jakob Erdmann‬‏ <‪namdre.sumo@xxxxxxxxx‬‏>:‬
you can download a test case that shows breakdown of flow at
The test case collects traffic data which can be used to plot a fundamental diagram with
sumo/tools/visualization/ edgedata.xml -i id -x density -y left -s --scatterplot --yfactor 60 --ylabel vehs/hour

It's pretty hard to observe spontaneous breakdown on a straight road because the disturbances travel upstream and take a while to develop into a full jam. For this reason the test uses 20km of road.


Am Fr., 1. Apr. 2022 um 14:02 Uhr schrieb dvir n <dvirer@xxxxxxxxx>:

I am trying to generate a realistic scene of a traffic break-down which is followed by many dynamic moving shokwaves in various lengths and speeds. 
In order to reach that road condition, I have modeled a simple 2 lane 4 km long straight road and inserted a high traffic flow (more than 2000 veh/hr/lane). However, no real serious shokwaves were observed. 
I was trying different car-following models which SUMO offers and edited their and the general Vtype attributes values. I have even decreased simulation time-step to 0.001 sec but still didn't succeed although high traffic values were observed (for example: Qmax = 2250veh/hr/lane, Km = 24veh/km, Vm = 94km/hr ). 

I will be glad to hear any suggestions for solving this problem and observing shockwaves which are caused just by the high density of vehicles in the road and not by an artificial interference such as adding ramp-in / generating a bottleneck / changing one car's speed or the road's shape / inserting non realistic Vtype's attribute values etc. 




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

Attachment: plot.png
Description: PNG image

Back to the top