Hello,
- DFROUTER assumes stable flows which means you have to aggregate the flow
values to overcome the temporal effects of traffic lights (alternatively,
you could structure you flow input data to correspond to the exact traffic
light phases).
- DFROUTER requires source and sink detections. When you say you have stop
bar data from a single intersection it sounds as if you do not have sink
detections on all sides. Is that so?
- do all your turning movements come from dedicated lanes? Otherwise the
flow solution is generally not unique
- you can always try flowrouter.py which was developed to overcome
limitations of DFROUTER (namly, guessing flow values where not all source
and sink values are known)

regards,
Jakob

Am Mo., 5. Nov. 2018 um 18:57 Uhr schrieb Flo AI <[email protected]>:

> Problem: We are generating a simulation from real-world detector data
> using DFROUTER but are facing large inaccuracies in the simulation for a
> single intersection.
>
> Situation: We're currently trying to accurately simulate traffic patterns
> on a 5-intersection corridor over a 24-hour time period.
> As an intermediary test step, we are simulating a single intersection
> using DFROUTER. Currently we are experiencing large variations in count
> accuracy on an hourly basis. We are seeing more inaccuracies for turning
> movement detectors than we are for thru-movement detectors but there are
> both over-estimates and under-estimates.
>
> Inputs: Lane-by-lane stop bar detector data. We have count data for every
> single approach lane across a 24-hour period. Resolution of 0.1s.
>
> If you have experienced these types of problems using DFROUTER and know of
> any fixes or potential sources of the error please let us know. Also, if
> you've been able to use detector data to drive simulations in SUMO for
> large corridor networks please let us know, we'd be grateful for
> advice/support.
>
> Best,
> Jatish
> _______________________________________________
> sumo-user mailing list
> [email protected]
> To change your delivery options, retrieve your password, or unsubscribe
> from this list, visit
> https://www.eclipse.org/mailman/listinfo/sumo-user
>
_______________________________________________
sumo-user mailing list
[email protected]
To change your delivery options, retrieve your password, or unsubscribe from 
this list, visit
https://www.eclipse.org/mailman/listinfo/sumo-user

Reply via email to