Hi Jakob,
thanks for your explanation so far! Attached you find the zipped sample-scenario (change ending from “.piz” to “.zip” to open it). Based on your answer I try
to explain my problem more detailed again:
Primary, my aim is to analyse how different attributes within the traffic simulation are distributed in case of stochastic traffic. Therefore I built up a more
complex simulation which shall approximate a stochastic traffic, which could occur in reality. In the second attachment you find the lanechange-output of the corresponding simulation. To analyse only the “correct” and “successful” lane changes, for now I
thought, that if the “changestarted” and “changeEnded” options are activated, only lane changes from vehicles in the output-file where these three change-states occur for an specific lane change are the successful ones and therefore are the only lane changes
I can investigate afterwards further.
Then I saw in the sample scenario, that a seeming “successful” lane change can occur in the corresponding lanechange-output with only the “changeStarted” and
“change” event. Due to this fact I am unsure under which condition I have to count a lanechange as “successful” in the more complex simulation, because, like mentioned above, sometimes all three states occur, sometimes not.
In addition, maybe it can give you a hint that there is never the reason “sublane” for lane changing in both, the sample and the complex simulation. Maybe this
could be caused by the fact that I chose a high value for the lateral resolution, because I want the sublane-Model to be activated for its additional options but I do not want to have real sublanes within the simulation. Can you give an answer how to achieve
this in general? I have in mind the following explanation of the doc:
Lane-changing model for sublane-simulation (used
by default when setting option --lateral-resolution
<FLOAT>). This model can only be used with the sublane-extension.
Caution:
This model may technically be used without activating sublane-simulation but this usage has not been fully tested and may not work as expected.
Thanks for your effort again.
Best regards
Nico
Von: sumo-user-bounces@xxxxxxxxxxx [mailto:sumo-user-bounces@xxxxxxxxxxx]
Im Auftrag von Jakob Erdmann
Gesendet: Mittwoch, 19. Juni 2019 14:05
An: Sumo project User discussions <sumo-user@xxxxxxxxxxx>
Betreff: [EXTERNAL] Re: [sumo-user] Definition/Condition "changeStarted"/"changeEnded" event in Lanechange Output
The cause of the observed behavior may be that a lane change is started for non-sublane reasons and then the reason is changed to 'sublane' before finishing the maneuver. This could indicate a bug. Please provide a minimal sample scenario.
Dear all,
i am currently writing my Master thesis regarding the safety validation of highly automated driving functions through simulative approaches. Therefore, amongst others, I use SUMO
as tool for the traffic simulation. In detail, I actually investigate the lane change behavior. Within the Sumo user doc (https://sumo.dlr.de/wiki/Simulation/Output/Lanechange)
the definition of the “change” element of the lane change output is explained, but not those of the “changestarted” events and “changeEnded” events in case of activating this option. Could you please explain what conditions have to be fulfilled to trigger
these events? The background is, that in my more complex simulation I saw that all three types of change events occur. But within a simple dummy simulation, where only some lane changes are performed, only the “changestarted” events and “change” events occur,
although vehicles perform “complete” lane changes. If necessary it is no problem to send you the corresponding files of the dummy simulation.
Thanks in advance.
Best regards
Nico
_______________________________________________
sumo-user mailing list
sumo-user@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/sumo-user