Hi Jakob,
Thank you for your response!
I'm not sure that I am fully clear on what you mean by "the used values are designed to filter out changes that could not be completed in time".
I interpret that to mean that normally any lane changes that have not been completed in the set time (in my example scenario 5 seconds) are completed by sliding. Setting lcMaxSpeedLatStanding to zero then filters out any lane changes that can only be completed
by sliding i.e. that lane changing decisions are made more reluctantly. Is that correct? I have done some playing around with the lanechange-duration value in my scenario (between 5-15) but didn't observe any significant changes in behaviour.
I am still working on a scenario reproducing the deadlocking behaviour with the sublane model. Of course when I want to show it, it doesn't want
to manifest :-)
Kind regards,
Gergely
From: sumo-user <sumo-user-bounces@xxxxxxxxxxx>
On Behalf Of Jakob Erdmann via sumo-user
Sent: Wednesday, June 19, 2024 5:29 PM
To: Sumo project User discussions <sumo-user@xxxxxxxxxxx>
Cc: Jakob Erdmann <namdre.sumo@xxxxxxxxx>
Subject: Re: [sumo-user] Is it possible to turn off sliding in the simple continuous lanechange model?
1. the simple continuous lc model does make use of lcMaxSpeedLatStanding, lcMaxSpeedLatFactor but not of lcMaxDistLatStanding. The documentation has been corrected, thanks!
The used values are designed to filter out changes that could not be completed in time but apparently don't work in all cases. I'll look into the matter.
2. the irregular queuing in the sublane model can be somewhat mitigated by setting lcLaneDiscipline="1" by at least in your scenario it doesn't prevent it entirely
3. If you can provide a scenario with deadlocks in the sublane model I would look into that as well
Dear SUMO community,
I see some conflicting information in the documentation on the parameters controlling sliding. Here (https://sumo.dlr.de/docs/Definition_of_Vehicles%2C_Vehicle_Types%2C_and_Routes.html#lane-changing_models)
the parameters lcMaxSpeedLatStanding, lcMaxSpeedLatFactor & lcMaxDistLatStanding are mentioned as belonging to both LC2013 and SL2015. However on the sublane page (https://sumo.dlr.de/docs/Simulation/SublaneModel.html#model_details)
lcMaxSpeedLatStanding & lcMaxSpeedLatFactor are mentioned as being additional parameters provided by SL2015. Do these variables in principle control both lane changing models or only the sublane model?
I have been trying to use lcMaxSpeedLatStanding & lcMaxDistLatStanding to eliminate sliding in the simple continuous lane change model (LC2013). We use that instead of the sublane
model because the latter is prone to deadlocks without sliding and exhibits undesirable queueing behaviour (screenshot attached) in congestion and we expect to see congestion in our simulation. It is important for our purposes that the vehicles behave realistically
as they would be observed from a car sitting in traffic i.e. they queue within lanes and do not slide sideways absent forward motion.
These parameters seem to not have an effect when using the simple continuous lane change model. Is this intended behaviour which the documentation does not reflect, is this a bug or am I not going about eliminating sliding correctly? Is it possible to eliminate
sliding in the simple continuous lane changing model?
I have attached a minimal repro scenario in which I am observing the sliding behaviour even though as far as I can tell it should be eliminated. I have attached some plots I made of the motion of some of the vehicles that show that there is transverse motion
when there is zero longitudinal motion.
Kind regards,
Gergely
This email and any files communicated with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error, please inform
us and delete this message. It is forbidden to copy, forward, or in any way reveal the contents of this message to anyone.
MXTREALITY UK LTD. trading as "MXT" | A company registered in England and Wales | Company number: 10431348 | The Courtyard, 4 Evelyn Road, London, United Kingdom, W4 5JL
_______________________________________________
sumo-user mailing list
sumo-user@xxxxxxxxxxx
To unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/sumo-user
This email and any files communicated with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error, please inform us and
delete this message. It is forbidden to copy, forward, or in any way reveal the contents of this message to anyone.
MXTREALITY UK LTD. trading as "MXT" | A company registered in England and Wales | Company number: 10431348 | The Courtyard, 4 Evelyn Road, London, United Kingdom, W4 5JL
|
