[sumo-user] Set up road-net supporting latitude and longitude

[Dayuan Tan]

Hi community,

Hope you everything is well!

I am wondering how to set up a road-net supporting latitude and longitude? I 
used a city osm file downloaded form open street map. I hope to make events 
(like a crash, or a vehicle starts moving) happen at a specific latitude and 
longitude. Thank you very much! Any reply would be appreciated very much!

Have a wonderful day!

Sincerely,
Dayuan
  





Dayuan Tan
UMBC



___
sumo-user mailing list
sumo-user@eclipse.org
To unsubscribe from this list, visit 
https://www.eclipse.org/mailman/listinfo/sumo-user

Re: [sumo-user] Cooperative lane changing in SL2015

[Hriday Sanghvi via sumo-user]

Hello again,

Thank you for the response. I understand that these 3 parameters are the
main factors that affect cooperative changes.

In a 2-lane scenario and a 3-lane scenario (road length = 1000m) with the
rightmost lane blocked by a stopped vehicle of length 300m at position
750m, there doesn't seem to be a big difference when setting
lcCooperative=1.0 vs. lcCooperative=0.0 for 100 vehicles to pass through at
a rate of 1 vehicles/hr -> only a difference of 4-8 simulation time
steps (seconds in my case). Wondering in what scenario this difference in
parameters would be most visible - if any clue?

Thank you.

Sincerely,
Hriday


On Mon, 14 Mar 2022 at 14:08, Jakob Erdmann  wrote:

> Hello,
> cooperative behavior is actually guided by three parameters that can be
> controlled separately:
> - lcCooperative: whether to perform a cooperative lane change if decreases
> ego speed, See Line 1375
> - lcCooperativeSpeed (defaults to lcCooperative): whether to adapt the ego
> speed to help other vehicles with lane changing. This is the code you
> already found
> - lcCooperativeRoundabout (defaults to lcCooperative): whether to use the
> inside lane on a multi-lane roundabout. See Line 1230
>
> lcCooperativeSpeed (myCooperativeSpeed) actually controls speed adaptions
> for the vehicle that wants to merge as well as the vehicle that wants to
> help (there is a known problem with this:
> https://github.com/eclipse/sumo/issues/9473)
>
> The 'getSlowest' code is for preventing overtaking on the right where this
> is forbidden. This is unrelated to cooperative behavior.
> To understand the speed adaptations that are advised
> (myLCAccelerationAdvices), activate the 'DEBUG_INFORM' flag in line 95 and
> then select vehicles in the simulation. (debug console output requires the
> debug build when on windows).
>
> regards,
> Jakob
>
>
> Am So., 13. März 2022 um 12:14 Uhr schrieb Hriday Sanghvi via sumo-user <
> sumo-user@eclipse.org>:
>
>> Hello,
>>
>> I am trying to understand the cooperative actions taken by vehicles in
>> SL2015 LC model. I am going through the C++ code and the start of this
>> line:
>> https://github.com/eclipse/sumo/blob/main/src/microsim/lcmodels/MSLCM_SL2015.cpp#L366
>> seems to have the crux of the cooperative actions taken.
>>
>>
>> https://github.com/eclipse/sumo/blob/21d11b1a26c4b781929b64416084e02a5f974cbe/src/microsim/lcmodels/MSLCM_SL2015.cpp#L370
>> seems to be the selection of speed for the vehicle that is cooperating to
>> let another vehicle onto its lane? OR is the selection of speed for the
>> vehicle that is changing lane into another one? - Not sure how *wanted* 
>> (variable)
>> is calculated.
>>
>>
>> https://github.com/eclipse/sumo/blob/21d11b1a26c4b781929b64416084e02a5f974cbe/src/microsim/lcmodels/MSLCM_SL2015.cpp#L2703
>> tells me that the speed above is also in some way related to checking the
>> speed of the slowest leaders on the neighbouring lanes?
>>
>> Please advise on where I could read more to understand in-depth - both
>> the lane changing and the speed change that is required by the cooperative
>> action? (For example, if I set the lcCooperative parameter to 1.0 (max))
>>
>> Thank you.
>>
>> Sincerely,
>> Hriday
>> ___
>> sumo-user mailing list
>> sumo-user@eclipse.org
>> To unsubscribe from this list, visit
>> https://www.eclipse.org/mailman/listinfo/sumo-user
>>
> ___
> sumo-user mailing list
> sumo-user@eclipse.org
> To unsubscribe from this list, visit
> https://www.eclipse.org/mailman/listinfo/sumo-user
>
___
sumo-user mailing list
sumo-user@eclipse.org
To unsubscribe from this list, visit 
https://www.eclipse.org/mailman/listinfo/sumo-user

Re: [sumo-user] Reinforcement learning for public transport system

[Lara Codeca]

Hi,

You can give a look to https://github.com/lcodeca/rllibsumoutils and
https://www.researchgate.net/publication/354696888_Using_Deep_Reinforcement_Learning_to_Coordinate_Multi-Modal_Journey_Planning_with_Limited_Transportation_Capacity,
presented last year at the SUMO User conference.

Best,
Lara

On Tue 15 Mar 2022 at 09:23, Pearla Hajjar  wrote:

> Hello,
>
> Thank you for your reply. I checked the link you sent, but I didn't see
> how it can help me with RL in sumo. I have managed to mess around using
> traci, but I don't get the idea of how RL will work with sumo.
>
> Every example I see is about the traffic lights only.
>
> Regards,
>
> Perla HAJJAR
>
>
> --
> *From:* sumo-user  on behalf of Lokesh Das
> 
> *Sent:* Sunday, March 13, 2022 5:45:32 AM
> *To:* Sumo project User discussions 
> *Subject:* Re: [sumo-user] Reinforcement learning for public transport
> system
>
> Hi
> The following link would be a very helpful and good starting point. SUMO
> works as an environment and TraCI has APIs to bridge between RL algorithm
> and SUMO.
>
> https://sumo.dlr.de/docs/TraCI.html
>
> Thanks,
> Lokesh
>
> On Fri, Mar 11, 2022 at 10:20 AM Pearla Hajjar 
> wrote:
>
> Hello all,
>
> I am new to sumo and I am currently creating a public transport system
> simulation. I want want to apply reinforcement learning on the buses to
> apply some strategy.
>
> I don't know where to start from regarding how to implement the
> environment of RL in sumo, how to link it to the graphical interface and
> how to store states.
>
> Is there any link or documentation on how to implement these parameters? I
> would appreciate any resources you provide.
>
> Regards,
>
> Perla HAJJAR
>
> Ph.D Student- Universite de Versailles SQ and Communauté d’Agglomération
> de Saint Quentin en Yvelines
>
> Direction des Mobilités- Navettes autonomes
> ___
> sumo-user mailing list
> sumo-user@eclipse.org
> To unsubscribe from this list, visit
> https://www.eclipse.org/mailman/listinfo/sumo-user
>
> ___
> sumo-user mailing list
> sumo-user@eclipse.org
> To unsubscribe from this list, visit
> https://www.eclipse.org/mailman/listinfo/sumo-user
>
___
sumo-user mailing list
sumo-user@eclipse.org
To unsubscribe from this list, visit 
https://www.eclipse.org/mailman/listinfo/sumo-user

Re: [sumo-user] Reinforcement learning for public transport system

[Pearla Hajjar]

Hello,

Thank you for your reply. I checked the link you sent, but I didn't see how it 
can help me with RL in sumo. I have managed to mess around using traci, but I 
don't get the idea of how RL will work with sumo.

Every example I see is about the traffic lights only.

Regards,


Perla HAJJAR



From: sumo-user  on behalf of Lokesh Das 

Sent: Sunday, March 13, 2022 5:45:32 AM
To: Sumo project User discussions 
Subject: Re: [sumo-user] Reinforcement learning for public transport system

Hi
The following link would be a very helpful and good starting point. SUMO works 
as an environment and TraCI has APIs to bridge between RL algorithm and SUMO.

https://sumo.dlr.de/docs/TraCI.html

Thanks,
Lokesh

On Fri, Mar 11, 2022 at 10:20 AM Pearla Hajjar 
mailto:pearlahaj...@outlook.com>> wrote:
Hello all,

I am new to sumo and I am currently creating a public transport system 
simulation. I want want to apply reinforcement learning on the buses to apply 
some strategy.

I don't know where to start from regarding how to implement the environment of 
RL in sumo, how to link it to the graphical interface and how to store states.

Is there any link or documentation on how to implement these parameters? I 
would appreciate any resources you provide.

Regards,


Perla HAJJAR

Ph.D Student- Universite de Versailles SQ and Communauté d’Agglomération de 
Saint Quentin en Yvelines

Direction des Mobilités- Navettes autonomes

___
sumo-user mailing list
sumo-user@eclipse.org
To unsubscribe from this list, visit 
https://www.eclipse.org/mailman/listinfo/sumo-user
___
sumo-user mailing list
sumo-user@eclipse.org
To unsubscribe from this list, visit 
https://www.eclipse.org/mailman/listinfo/sumo-user