Re: [sumo-user] junction inspect with "show connections over junctions": not understood case of priority flow (with TLS)
Jakob, I found a solution for the problem through your doc link to right of way. After changing rightOfWay from Default to edgePriority, the priorities show up as desired. I also have a definitive answer, that this is the root of all so called request entries with any xml junction description.
However, the default should do it as well, as the attribute "speed" is the same (except for the oneway edge from west which is lower). Most junctions have Default rightOfWay, and are as intended.
Just while writing this, I 'discovered' that I can inspect the connections characteristics not via edit connections but simply via "inspect" mode. Until now I only used the inspect mode not only for *junction* inspection, but also for deletion of connections (e.g. many 180° turns that are generated from OSM import). Just now I'm aware that clicking on a connect line goes associated with showing a lot of parameters in the left pane. Is this the way to customize special connect priorities?
All of this seems quite important in order to obtain a meaningful default tls configuation, and indeed, your doc reference tells:
> Right-of-way computation also influences connection-guessing and the generated traffic light program
Am 03.02.2022, 23:02 Uhr, schrieb Jakob Erdmann <namdre.sumo@xxxxxxxxx>:
The right-of-way rules are initialized from the edge 'priority' values. Lowering the east-west priorities or increasing the north-south priorities should achieve the desired effect.
In all junction cases in my net with TLS (and also without), the cyan resp. white connection lines are drawn according to the "logical" prioity. In the case of the attachment, this is not the case (one screenshot in T- and one in I-mode).
I'd suggest that the edge priorities define the correct default priority connection scheme.
settings clockwise, from the north edge: 10, 9, 12, 3.
How to establish a correct representation as desired?_______________________________________________
sumo-user mailing list
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user