Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [sumo-user] Different vehicle behavior in same surroundings

Hello,
there is currently no way to reset the random number generator except by restarting the simulation.
So running scenario A,B will have different behavior in B compared to running B alone.
A solution would be in https://github.com/eclipse/sumo/issues/1418 (and then saving state before running B)
regards,
Jakob

Am Fr., 17. Apr. 2020 um 11:32 Uhr schrieb Branka Mirchevska <brankamircevska@xxxxxxxxx>:
Hello Jakob,

I took a look at the output files, there has been an oversight and I wasn't
using the randomly generated seed, thanks for pointing that out!

I would like to know, if there is a way I can make sure that whenever I ran
scenario B, regardless of how many scenarios I have ran before it (without
restarting Sumo), I get exactly the same behavior of the surrounding
vehicles (given of course that I don't change the behavior of the one
remotely controlled vehicle).

This is important because I want to keep track of collisions caused by the
remotely controlled vehicle, and a slight change of the behavior of the
surrounding vehicles, next time I ran the simulation, may result in no
collision.

This is the setting: I have fixed scenarios A, B, C, D, ...., Z (different
num. of vehicles, starting positions etc). I automatically evaluate the
performance of the remotely controlled vehicle on all of them. After the
evaluation is done, I detect that in scenario K there has been a collision.
I want to directly run just that one scenario and inspect it to see what
happened, without the possibility that I see something else now due to
slightly changed behavior of the other vehicles.

Thank you!

VG
Branka



--
Sent from: http://sumo-user-mailing-list.90755.n8.nabble.com/
_______________________________________________
sumo-user mailing list
sumo-user@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user

Back to the top