Hello Everyone I want to ask about priority rule in a two-lane roundabout, I followed the guide step by step but still, there are some crashes between entering vehicles and circulating vehicles during the simulation. I tried different marks positions but still, the situation is the same. I will be extremely appreciated if you could help me hashtag#simulation hashtag#vissim hashtag#roundabout
I think it is very difficult to have continuous movement which is the main advantage of roundabouts for two lane roads...It would be interesting to see the solution for that...thanks for sharing such a question.
Can you show a picture or the inpx file?
I do not have Vissim right at hand, so a picture of the roundabout and the attributes of the priority rules would be useful.
Some time ago I succeded in simulating roundabouts when putting also some stop lines into the roundabout to avoid crashing with vehicles that are already entered the roundabout from the approaches.
But this will influence the concept of priority rule that vehicle at the Circulating lane have priority over entering vehicle
Some ideas from the picture
- Do I understand that correctly and you use multiple priority rules with lots of conflict markers? That is not in itself an error, just a bit awkward. A priority rule can have multiple conflict markers assigned, so I would expect one priority rule per approaching lane and two conflict markers per conflicting roundabout lane: One with a space gap to ensure no vehicles are in the conflict area, and one with a time gap to ensure nothing approaches from the roundabout.
- The way the roundabout is modeled, the vehicles in the roundabout do not care what is coming. So there is a time gap between the [front of the] approaching vehicle passing the priority rule and the same being on a roundabout link where neither the approaching vehicle nor the vehicle inside the roundabout know about each other (they are on completely different links, even though these seem to overlap, hence they could just as well be on different ends of the network). The time/space gaps must be set to values that ensure that there is no collision during this time gap. That requires some fiddling. If it happens that you need unrealistic settings to avoid collisions, you have reached the end of what you can model with priority rules. Either you accept "collisions" then or unrealistic parameters.
- Something that confuses me to no end: Why do you use priority rules at all? Except for special cases, conflict areas are usually better. Priority rules are basically stop and go (if a conflict marker registers an incoming vehicle, they act like a red traffic light, if nothing is registered, they are practically invisible). Conflict areas, however, also do some guessing of vehicle trajectories, and if a suitable gap is found, vehicles on the approaching link here would try to adjust their driving to fit into the gap. Plus a vehicle on the conflict area is always handled as an obstruction, even if the vehicle has no right of way, so collisions are much harder to create.
- You could try out some debugging on your own if you have not done so yet. Mark one of the near coincidences in time (say, at 105 s). Then do the simulation again and stop a bit before, say, 100 s. Open the list of vehicles in the network, and show the attributes: "Driving state", "Interaction target", "Interaction type" and "Interaction number". Then look at the vehicle approaching the roundabout and continue with single-stepping to understand how it reacts.
In the figure that you have provided, there are multiple number of priority rule markings. Is it because the you had fixed separate rules for individual vehicle types? And also for each of the circulating lanes, i guess? If so, how did you fix the values for gap time and headway. Increasing these values might help you to avoid these collisions. But again, on what basis or to what extent will you increase the values of gap time and headway.
As Ulf Lorenz had suggested, I also would have preferred in using conflict areas over priority rules for your case.
Ulf Lorenz Mithun Mohan thanks a lot for the informative information, I will try what you suggest, and I will inform you the result
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators