| Re: [sumo-user] Fwd: Merging and vehicle behavior |
It's a complicated question:- on linux, there is hardly a difference between running with sumo-gui and delay 0 versus sumo- on windows, delay 0 introduces more delay than it should most likely due to some event handling overhead- on windows, setting delay 10 has a stronger slowdown effect than on linux- nothing was changed between 1.15 and 1.17 that should have changed this- bugfixes in sumo itself can lead to changing scenario behavior and even more/less jamming which affects the number of vehicles at any given time and thus simulation speedSo to figure this out:- is the speed difference maybe from sumo itself? is there a speed difference in your scenario when running without gui using 1.15 and 1.17?- did the number of UPS (updates per second reported in the verbose log) change when running your scenario with either version (again without gui)?- what OS are you running your tests on?- what are the running times for sumo vs sumo-gui with delay 0 and what is the reported real time factor and UPS respectively?_______________________________________________Am Sa., 13. Mai 2023 um 05:17 Uhr schrieb Albert Jeans <ajeans2@xxxxxxxxx>:Hi Jakob,Thanks for all the tips for improving my simulations!One thing I've noticed recently, and I don't know if it started when I upgraded to 1.17 from 1.15, but the simulations seem to be running much more slowly when the Delay is set to 0 ms. In fact there doesn't seem to be any difference in speed for delays about 200 ms or less. I used to be able to run the simulation very quickly to get to a certain time. Is there some other setting which is limiting how fast the simulation runs?Thanks,Albert Jeans_______________________________________________On Fri, May 12, 2023 at 1:52 AM Jakob Erdmann <namdre.sumo@xxxxxxxxx> wrote:some notes on the scenario:- you have emergency braking because 3s of yellow time are insufficient for safe braking if people are routinely over-speeding by 30% (as configured) on a road speed limit of 22.22s- you have some zipper junctions where you set individual links to pass=true. This is causing collisions because vehicles on these links do not follow the zipper behavior (instead assuming priority) - admittedly it does reduce required lane change and thus helps with traffic flow- you have vehicles that want to turn left and signaling right because they would prefer to be on the *other* left turn lane but in fact they can pass the intersection just fine and change downstream (the same for right turns)- it helps a lot to set lcAssertive to a higher value i.e. =2 to increase gap acceptance- there is a known problem with lane changing efficiency when vehicles have unsynchronized action steps. A simple remedy would be to set your step-length to 0.6 and not use action-step-length (https://github.com/eclipse/sumo/issues/4457)regards,Jakob_______________________________________________Am Do., 11. Mai 2023 um 16:21 Uhr schrieb Albert Jeans <ajeans2@xxxxxxxxx>:_______________________________________________Hi Jakob,Thanks for your reply. I'm attaching all the files (hopefully) that you need to run my scenario (minus the decals). The problem seems to occur when the two right turn lane queues get very long (please scale the traffic by 1.3 or higher).Like real drivers, some of the vehicles jump the queue, expecting to be able to merge farther down. Here's a screenshot of cars with their right turn signals on in a left turn lane.Many thanks for your unfailing support!Albert JeansMountain View, CA USAOn Wed, May 10, 2023 at 12:57 AM Jakob Erdmann <namdre.sumo@xxxxxxxxx> wrote:The strategic foresight extends to 3000m so drivers should use all the available distance for weaving. I did a quick test and put ~4500 veh/h down that road with 50% of cars forced to weave and didn't observe any problems. Can you provide your scenario files?_______________________________________________Am Mi., 10. Mai 2023 um 01:44 Uhr schrieb Albert Jeans <ajeans2@xxxxxxxxx>:I am trying to simulate a 4-lane freeway offramp where the left 2 lanes turn left and the right 2 lanes turn right at a signalized intersection at the end of the ramp. The ramp is about 500 m long, but traffic is often heavy enough that queues can fill the entire ramp. Traffic comes from 2, 2-lane highways and there is a significant amount of "weaving." Real drivers know which way they want to turn and merge into the appropriate lanes well before the traffic signal. SUMO vehicles, however, often seem to wait too long, and as a result, end up at the intersection in the wrong lanes which causes gridlock. I've tried putting additional zipper junctions within the onramp to "encourage" vehicles to get into their proper lanes but that still doesn't work all the time._______________________________________________How far in advance do SUMO vehicles start planning to change lanes ahead of an intersection and is there a way to adjust this parameter?By the way, this is a real road. You can see it on Google Maps at (N 37.410905090334374, -122.07430147214615).Thank you for any suggestions you might have!Albert Jeans
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
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
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
