Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [sumo-user] Questions about creating a capacity drop and lane-changing before a zipper junction

Hi Jakob,

Thank you for your guidance. I find them very useful.

But my first question seems not solved yet. I'm sorry that I did not describe the question clearly. The inflow did not come from the input route file, but instead from the edge data output results, which were declared in the additional file (the 'departed' value of this edge is 1246 in fifteen minutes), I also put a loop detector at the beginning of the road segment and the inflow of 4 lanes sums up to 4952veh/h. I just checked the GUI and found that congestion didn't spread to the input position before the end of the data collection cycle.

E1, E2, and E3 are from the upstream section to the downstream one.  E1 and E2 are 4km long (in total) and have 4 lanes, while E3 is 200m long and has 2 lanes. 

image.png
This is the output of loop detectors at the beginning of the road segment, respectively on each lane.

image.png
This is the edge data output file, which I collect the data from. The collection cycle is 15 min. The inflow is collected as the 'departed' at E1, and the throughput is collected from the 'entered' value of E3, i.e. 717 veh (15min,2lanes). I'm wondering if some parameters of vehicle type are set inappropriately. There are two vehicle types in my route file, which are different in driving styles:

<vType id="type_1" speedDev="0.5" accel="2.5" decel="4.0" tau="1.5" length="5" color="0,1,0" vClass="passenger" maxSpeed="20" carFollowModel="Krauss" laneChangeModel="LC2013" lcCooperative="0.2" lcSpeedGain="0.5" lcPushy="0.5"/>
<vType id="TYpe_2" speedDev="0.5" accel="2.5" decel="4.0" tau="1.8" length="5" color="0,1,1" vClass="passenger" maxSpeed="20" carFollowModel="Krauss" laneChangeModel="LC2013" lcCooperative="0.8" lcSpeedGain="0.2"/>

Looking forward to your reply. Thanks in advance.

Best regards,

Xuanmian He



On Mon, Jul 15, 2024 at 1:54 PM Jakob Erdmann via sumo-user <sumo-user@xxxxxxxxxxx> wrote:
From the low spacing in your screenshot, I'm guessing that you are measuring inflow by the number of vehicles specified in the input file rather than by inflow into the simulation.
Most likely, you have delayed departures due to jamming on the inflow edge and are only ever simulating inflow of about 3000 veh/h.

The easiest fix would be to do the following:
- increase the length of the inflow edge
- set departSpeed="avg"

Check for the "departDelay" value in tripInfo-output or statistics-output to make sure your vehicles are actually able to enter. 

The counterintuitive spacing comes from the transition between jammed inflow and then free-flow at the merging point.
You can tune the distance at which zipper behavior starts with connection attribute visibility distance: https://sumo.dlr.de/docs/Networks/PlainXML.html#explicitly_setting_which_edge_lane_is_connected_to_which

regards,
Jakob


Am Mo., 15. Juli 2024 um 04:56 Uhr schrieb Xuanmian He via sumo-user <sumo-user@xxxxxxxxxxx>:
Dear SUMO Team,

I'm building a lane-drop bottleneck scenario and trying to create a capacity drop. I tried the 'priority' and 'zipper' modes for the junction, but both of them didn't perform the capacity drop even when demand greatly surpassed the capacity. Also, I turned 'speedDev' value to 0.05 and 'lcCooperative' to very low(0.2), as taught in SUMO documentation, but it still didn't work.

I adjusted the input flow rate and did the experiment. The relationship between the inflow of the network and the throughput of the bottleneck is attached. In my scenario, the bottleneck has 2 lanes and the upstream section has 4 lanes. After inflow exceeded 4000veh/h (i.e. 2000 veh/lane/h), throughput was expected to drop but it didn't.
4e9c8231ada3f54f03a178a480a941d.png
Another question is about lane-changing behavior in front of a zipper junction. It seems that vehicles tend to keep a higher spacing for merging and start to accelerate about 100 meters in front of the merging point. How to make them wait and pass alternately only at the merging point?

image.png
I appreciate your help.

Best regards,

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