When the option --no-internal-links is enabled, streets are made
artificially longer so that the total route length matches the
geometrically traveled distance despite "jumping" over intersections. The
vehicle positions are interpolated to compensate (resulting in visually
short distances while que
Dear all,
found the culprit: the option 'no internal links' was switched on in
NETEDIT, I am unsure how that happened. I only found out by comparing my
network to a previous one I built.
The 'no internal links' option seems somewhat odd, as vehicles are not
displayed on their actual position.
Dear all,
in a network I made I am experiencing behaviour of vehicles that seems
odd. The vehicles drive normally, but upon reaching a junction, teleport
to the other side. Also, while waiting for a traffic light, they seem
not to leave any distance between them.
I am unaware of having chang
Sorry,
this is beyond the scope of this mailing list.
Take a look at individual simulations to determine why the target function
would react favourably to increased durations.
At high traffic demand it is a common strategy to increase the cycle length
as every switch reduces throughput (there are s
I am trying to optimize the traffic light timings based on reinforcement
learning ,but I meet a problem ,the performance index 'Duration' is
increasing explosively.At the beginning,the Duration is about 8092 ms,but
soon,it increased to 13 ms,and is still increasing.
I don't know what is wrong ,
--default.junctions.radius
is correct. It was a typo on my part.
Thanks,
Jakob
2018-06-26 8:59 GMT+02:00 Lara CODECA :
> Hello Jakob,
>
> I'm using the new netconvert on my old scenario :)
> If I use --default.junctions.size 5 I get:
>
> Error: On processing option '--default.junctions.size':
