Hi Jakob,
I've created the vehicles file running sumo with --vehrouter-output and now
I get this error:

Step #84600.00 (1534ms ~= 0.65*RT, ~9714.47UPS, vehicles TOT 49272 ACT
14902 BUF 571842)  ^M
Simulation ended at time: 84600.00
Reason: The final simulation step has been reached.
Performance:
 Duration: 60435.81s
 Real time factor: 1.39983
 UPS: 12824.598592
Vehicles:
 Inserted: 49272 (Loaded: 621114)
 Running: 14902
 Waiting: 571842
Teleports: 13175 (Collisions: 14, Jam: 9812, Yield: 2463, Wrong Lane: 886)

Begin time:0
Traceback (most recent call last):
  File "/home/u7262/sumo-1.9.2/tools/tlsCycleAdaptation.py", line 447, in
<module>
    main(options)
  File "/home/u7262/sumo-1.9.2/tools/tlsCycleAdaptation.py", line 427, in
main
    groupFlowsMap = optimizeGreenTime(tl, groupFlowsMap, phaseLaneIndexMap,
currentLength, options)
  File "/home/u7262/sumo-1.9.2/tools/tlsCycleAdaptation.py", line 342, in
optimizeGreenTime
    (critialFlowRateMap[i] / sum(critialFlowRateMap.values())) -
options.yellowtime + options.losttime
*ZeroDivisionError: float division by zero*

Why is this happening? After some code checking I've seen that critical
flow ratios (critialFlowRateMap) depends on: maxFlow, which is set to 0 in
the code; satFlows, which is equal to 3600. / options.satheadway;
and phaseLaneIndexMap which I think is related with green phases. There is
anything I can or should change to get better results?

As you can see, barely an 8% of the total vehicles are inserted into the
network, and I'm almost positive that it is because of the jams induced by *too
many red-phases* in most of the traffic light signals, specially at
intersections due to a poor coordination of traffic light signals. It is
tlsCycleAdaptation.py the right tool to fix this? If not, what are my
options?

Best regards,
Alejandro.

El vie, 24 sept 2021 a las 12:21, Jakob Erdmann (<[email protected]>)
escribió:

> Hello,
> the tool tlsCycleAdaptation.py expects as input a route file that contains
> vehicle elements (with embedded routes). These vehicles are counted to
> determine the traffic level at each controlled intersection.
> In your input, only route definitions are given but no vehicles. Hence no
> traffic information can be used.
> route files suitable for the script can be obtained from duarouter or when
> running sumo with option --vehroute-output
>
> regards,
> Jakob
>
> Am Mo., 20. Sept. 2021 um 10:18 Uhr schrieb Alejandro Rodríguez Sánchez <
> [email protected]>:
>
>> Hi everyone,
>> I'm trying to optimise my traffic simulations by
>> using tlsCycleAdaptation.py to adapt the tls cycles as I've seen that some
>> tls is in red phase for too long and causing some undesired jams.
>> However, the python script is skipping all the TLS and therefore no
>> optimization is done, as the message shows:
>>
>> Skipping TLS 'joinedS_13' due to unused states (116 states, 114
>> connections)
>> Skipping TLS 'joinedS_16' due to unused states (124 states, 121
>> connections)
>> Skipping TLS 'joinedS_21' due to unused states (30 states, 29 connections)
>> [...]
>> Skipping TLS 'joinedS_78' due to unused states (32 states, 31 connections)
>> Skipping TLS 'joinedS_94' due to unused states (23 states, 22 connections)
>>
>> *There are no flows at the given intersections. No green time
>> optimization is done.*
>>
>> I have seen that this occurs when there are more states that connections.
>> What does exactly mean? And there is something I can do to solve this and
>> successfully optimise the TLS cycles?
>>
>> I attach my network and routes file in case it is useful to clarify this:
>>
>> https://drive.google.com/file/d/153mRU0kQKiJV2K083LZ8psGDp0RD3xVk/view?usp=sharing
>>
>> Best regards,
>> Alejandro.
>> --
>> Alejandro Rodríguez Sánchez
>> PhD student
>> CIEMAT // Universidad Complutense de Madrid
>> 28040 Madrid
>> email: [email protected]
>> Website: https://alexrsanchez.github.io/en/
>> _______________________________________________
>> sumo-user mailing list
>> [email protected]
>> To unsubscribe from this list, visit
>> https://www.eclipse.org/mailman/listinfo/sumo-user
>>
> _______________________________________________
> sumo-user mailing list
> [email protected]
> To unsubscribe from this list, visit
> https://www.eclipse.org/mailman/listinfo/sumo-user
>


-- 
Alejandro Rodríguez Sánchez
PhD student
CIEMAT // Universidad Complutense de Madrid
28040 Madrid
email: [email protected]
Website: https://alexrsanchez.github.io/en/
_______________________________________________
sumo-user mailing list
[email protected]
To unsubscribe from this list, visit 
https://www.eclipse.org/mailman/listinfo/sumo-user

Reply via email to