Skip to main content
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
[sumo-user] Teleporting and failed lane changing at motorway fork

Hi all,

 

first of all,  thanks again for the helpful reply to my previous question – sorry to bother with a new one 😉

 

I’m in the calibration process of a motorway section. The network contains a fork where a three-lane roadway splits into two two-lane roadways. Through traffic must merge left at this fork.

 

Currently, I am using a step length of 0.1s with action step length of 1.0s. I am observing many stuck vehicles not able to perform the necessary lane change in time. For now, I am using the default car following and lane changing parameters (except for the speed acceptance, which follows a normal distribution for a passenger car and a truck type).

 

While I had few problems with failed lane changes with a simulation step of 1s, the severity of congestion due to stuck vehicles seems to increase as I reduce the step length (leaving action step length the same).

 

I have a few questions on this issue:

 

  1. Default value for SUMOs time-to-teleport.highways option

 

The doc states that when teleporting, highways are distinguished from other roads if its speed limit is > 69 m/s = ca. 250 km/h. Is it possible that this distinction was meant to be made at ca. 70 km/h instead? Is this wrong in the doc? Because if I use the following configuration, vehicles are removed quickly even with edge speed limits lower than 69m/s:

 

<time-to-teleport.highways>0.1</time-to-teleport.highways>

<time-to-teleport>300.0</time-to-teleport>

 

  1. What does the –time-to-teleport.highways default value of zero mean?

 

When I do not specify a value for –time-to-teleport.highways, i.e. with default value zero, I still see vehicles getting stuck and waiting. In contrast to a –time-to-teleport value of 0, teleporting doesn’t seem to be disabled at all, but it seems to match the –time to teleport value. Is that correct?

 

  1. Influence of simulation step length on teleporting and/or Lane Changing?

Is the teleporting and/or the lane changing behavior in any way influenced by the simulation step (given that action step length remains fixed)? As said, with a simulation step of 1.0s, failed lane changing is by far less frequent - I see only a few speed drops, but no stopping vehicles. All LC params, as well as tau and the other CF params, are at their default values in  both cases.

 

Thanks for a reply!

 

Kind regards

Jan


Virenfrei. www.avast.com

Back to the top