| Re: [sumo-user] Low flow in lane with flowrouter |
_______________________________________________Hi Jakob,
Those are some great suggestions, thanks. I see now looking through the sourcecode that there is an option --lane-based, does this mean that I could get the output per lane instead then? In regards to changing the aggregation time, will I get vehicles all coming in chucks every 5 minutes (or whatever number I choose) then?
BR,Fredrik
From: sumo-user-bounces@xxxxxxxxxxx <sumo-user-bounces@xxxxxxxxxxx> on behalf of Jakob Erdmann <namdre.sumo@xxxxxxxxx>
Sent: Tuesday, April 2, 2019 14:53
To: Sumo project User discussions
Subject: Re: [sumo-user] Low flow in lane with flowrouterFlowrouter only sums up the lanes and doesn't concern itseelf with flow per lane
So there are two seeprate issues to this:
1) total flow: Looking at time 90, the simulation has a deficit of ~1300 vehs/hour. This could be due to the low aggregegation interval of 1 minute because flowrouter does not take driving time between detectors into account. Whenever the flow changes a lot, flowrouter tends to underestimate total flow. To compensate, set a higher value for option -i
2) vehicle distribution across the lane. This distribution comes out of the simulation dynamics and depends mostly on
- configuration for insertion lanes and insertion speed
- the parameterization of the vehicle fleet in regard to lane changing
- the distribution of desired speeds
regards,Jakob
Am Di., 2. Apr. 2019 um 12:43 Uhr schrieb Fredrik Nilsson <frnilss@xxxxxxxxxxxxxxxxxxx>:
_______________________________________________I accidentally attached the Speed over time plot instead of the Flow over time plot, it's attached here.
From: sumo-user-bounces@xxxxxxxxxxx <sumo-user-bounces@xxxxxxxxxxx> on behalf of Fredrik Nilsson <frnilss@xxxxxxxxxxxxxxxxxxx>
Sent: Tuesday, April 2, 2019 12:39
To: Sumo project User discussions
Subject: [sumo-user] Low flow in lane with flowrouterHi,
I've been tuning a simulation to match real detector data and I have a somewhat odd issue that I was wondering if anyone can perhaps come up with a reason for. I have imported the network from OSM and made some slight changes, I have attached a picture of the network (blue lines are where detectors are present, one per lane) and the detectors whose results I want to match to reality are the ones inside the red circle (from the left to the right they are detectors 35138, 35139, 35140 and 35141).
I have also attached pictures of the flow over time for both real data and simulated data (red is simulation, blue is real data) and a density-flow fundamental diagram scatter plot (red is simulaton, blue is real data).
Now, as can be seen on 35138 (the left most lane in the red circle) the flow is very low compared to reality and this problem only seems to exist on that lane. I have generated the flow and route files with flowrouter.py with 1 minute intervals (the same as the real world data that it was generated from).
Can anyone think of a reason why my flow would be so low on this specific lane and if perhaps I could make some changes to flowrouter settings to make it better?
Sorry for the long question and best regards,
Fredrik
sumo-user mailing list
sumo-user@xxxxxxxxxxx
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
sumo-user@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/sumo-user