Hello, when testing different intersection configurations it is a common occurence that some scenarios do not permit all vehicles to arrive (or even depart) within a given time frame due to congestion. In that case you may either end the simulation early (and deal with the reduced number of completed trips in your evaluation) or you may wait till all vehicle have arrived (and deal with the increased duration in your evaluation).
- Sumo can automatically reduce the number of vehicles imported from a route file by setting the option --scale. - if a queue extends beyond the length of the E2 detector, that part of the queue is not registered. - tau is used in the car following models. see src/microsim/cfmodels (MSCFModel_Krauss.cpp if you didn't configure another model). - thank you for the suggestion regarding lane geometries. We intend to fix but did not yet get around to it: https://github.com/eclipse/sumo/issues/1758 regards, Jakob 2018-04-19 21:54 GMT+02:00 Rahimi, Tariq Rahim < tariqrahim.rah...@rockets.utoledo.edu>: > I appreciate your time Jakob, > > > I was using the version 32 and it works fine now in the latest development > version. > > > I want to simulate 1 hour worth of traffic of 4933 vehicles, I am > generating the route_file through python so that the directions and > vehTypes can be random. > > I still have problem with depart, if I change it to 3600 timesteps as > well, all the vehicles won't finish in the 60 minutes allotted. > > What is a good solution for this through python? I tried --step-length > when running the configuration file in SUMO-GUI, it didn't work. Can SUMO > reduce the number of vehicles by percentage if I specify 40 minutes or 90 > minutes? I am using probabilities here in python as it is not equally > spaced if I am not wrong. > > > Also, you said I can obtain queue-length from E2 detector, what variables > can I get it from (adding or subtracting). If I use the E-2 detector for > just 50 meters and the queue length goes above that, what's the case here? > > > Additionally, can you give me the reference for tau's code file, I want to > know what it triggers? > > > Just a suggestion, can you guys think of a way so that the a 4-leg > intersection can have 2 outbounds and 4 inbounds (2 left turns) and the > geometry should align themselves automatically (the left storage lane would > be in front of the opposite directions left storage lane, currently the > medians are aligned), If I have a scenario with 4 inbounds and 2 outbounds, > it shows crashes. > > > Thank you, > > > * ------------------------------------- Tariq R Rahimi* > Graduate Student > Department of Civil Engineering > University of Toledo > Toledo, OH 43606 > > O: NI 1076 > P: 567-806-0945 > E: tariqrahim.rah...@rockets.utoledo.edu <http://www.utoledo.edu> > tariqrrah...@gmail.com <http://www.utoledo.edu> > > ------------------------------ > *From:* Jakob Erdmann <namdre.s...@googlemail.com> > *Sent:* Wednesday, April 11, 2018 2:19:36 PM > > *To:* Rahimi, Tariq Rahim > *Cc:* Sumo project User discussions > *Subject:* Re: A question about timestep and tau > > Did you install version 0.32.0 or did you use the latest development > version from here: http://sumo.dlr.de/wiki/Downloads#SUMO_-_Latest_ > Development_Version > <https://urldefense.proofpoint.com/v2/url?u=http-3A__sumo.dlr.de_wiki_Downloads-23SUMO-5F-2D-5FLatest-5FDevelopment-5FVersion&d=DwMFaQ&c=heEcP2AsrLOv4XP8e7I9NA&r=hcDhabDH71qjaXA3guxceND6Pl-jLmH9TuxEWZxKw1xUdKLCB6do1p6Q6bjgl8Fe&m=s6pSl2G1DvLPFuwCOHqBbvrfyOhyrKvjZx2XHd8COEw&s=ai8TcAqafGE6EuPFlniJEBMtwz8XBZoJ_P-ja_1fGYE&e=> > ? > > You can obtain queue-length using the lanearea detector: > http://sumo.dlr.de/wiki/Simulation/Output/Lanearea_Detectors_(E2) > <https://urldefense.proofpoint.com/v2/url?u=http-3A__sumo.dlr.de_wiki_Simulation_Output_Lanearea-5FDetectors-5F-28E2-29&d=DwMFaQ&c=heEcP2AsrLOv4XP8e7I9NA&r=hcDhabDH71qjaXA3guxceND6Pl-jLmH9TuxEWZxKw1xUdKLCB6do1p6Q6bjgl8Fe&m=s6pSl2G1DvLPFuwCOHqBbvrfyOhyrKvjZx2XHd8COEw&s=X_RGJRoWtLQIf_oFtEsWSIUIsSkKTy_SYRHH6TH0mBQ&e=> > > A good measure of congestion is time loss. A global value can be obtained > by activating the statistics output: http://sumo.dlr.de/wiki/ > Simulation/Output#Aggregated_Traffic_Measures > <https://urldefense.proofpoint.com/v2/url?u=http-3A__sumo.dlr.de_wiki_Simulation_Output-23Aggregated-5FTraffic-5FMeasures&d=DwMFaQ&c=heEcP2AsrLOv4XP8e7I9NA&r=hcDhabDH71qjaXA3guxceND6Pl-jLmH9TuxEWZxKw1xUdKLCB6do1p6Q6bjgl8Fe&m=s6pSl2G1DvLPFuwCOHqBbvrfyOhyrKvjZx2XHd8COEw&s=e6v--ISk2YnpK2wqQVUtwqGAZj2KEjck_yJvrBo1bqE&e=> > > regards, > Jakob > > 2018-04-11 19:33 GMT+02:00 Rahimi, Tariq Rahim <TariqRahim.Rahimi@rockets. > utoledo.edu>: > > I appreciate your response, > > > I installed the newer version just yesterday and I am still getting > waitSteps in the Tripinfo output and I don't get any waitingTime. > > > Additionally, where can I get a good measure of queue length and > congestion? > > > Thank you, > > > * ------------------------------------- Tariq R Rahimi* > > ------------------------------ > *From:* Jakob Erdmann <namdre.s...@googlemail.com> > *Sent:* Tuesday, April 10, 2018 3:00:19 AM > *To:* Rahimi, Tariq Rahim > *Cc:* Sumo project User discussions > *Subject:* Re: A question about timestep and tau > > Hello, > The value of waitSteps counts the number of simulation steps in which a > vehicle was waiting. So if your vehicle waits for 3 seconds and you > simulation with --step-length 0.2 that would be 15 waitSteps. > In the latest development version, the attribute waitSteps is replaced > with waitingTime which is measured in seconds. ( > http://sumo.dlr.de/wiki/ChangeLog#Git_master > <https://urldefense.proofpoint.com/v2/url?u=http-3A__sumo.dlr.de_wiki_ChangeLog-23Git-5Fmaster&d=DwMFaQ&c=heEcP2AsrLOv4XP8e7I9NA&r=hcDhabDH71qjaXA3guxceND6Pl-jLmH9TuxEWZxKw1xUdKLCB6do1p6Q6bjgl8Fe&m=WFYrbGGcPNeSYHOh2KWsVWWO3Zdu-Yc8HIH8OxM9TG4&s=pxZRu4YuCEwEuwWk0S20Sc4c4N1nUWfSyE6vtpXjt7Q&e=> > ). > > The value of tau determines the desired time-headway between successive > vehicles. The exact usage of this value depends on the car following model > (see http://sumo.dlr.de/wiki/Definition_of_Vehicles,_Vehicle_ > Types,_and_Routes#Car-Following_Models > <https://urldefense.proofpoint.com/v2/url?u=http-3A__sumo.dlr.de_wiki_Definition-5Fof-5FVehicles-2C-5FVehicle-5FTypes-2C-5Fand-5FRoutes-23Car-2DFollowing-5FModels&d=DwMFaQ&c=heEcP2AsrLOv4XP8e7I9NA&r=hcDhabDH71qjaXA3guxceND6Pl-jLmH9TuxEWZxKw1xUdKLCB6do1p6Q6bjgl8Fe&m=WFYrbGGcPNeSYHOh2KWsVWWO3Zdu-Yc8HIH8OxM9TG4&s=ReunekgtCFa-RhcSIXd0kvHxyEyDHtxhdFyogahxqlU&e=> > ) > > To model different perception-reaction times, see > http://sumo.dlr.de/wiki/Car-Following-Models#Model_Parameters > <https://urldefense.proofpoint.com/v2/url?u=http-3A__sumo.dlr.de_wiki_Car-2DFollowing-2DModels-23Model-5FParameters&d=DwMFaQ&c=heEcP2AsrLOv4XP8e7I9NA&r=hcDhabDH71qjaXA3guxceND6Pl-jLmH9TuxEWZxKw1xUdKLCB6do1p6Q6bjgl8Fe&m=WFYrbGGcPNeSYHOh2KWsVWWO3Zdu-Yc8HIH8OxM9TG4&s=BL5SFhMbMbhdbrBaFVX7lFagcvGK1aOtCHt_MpQ2RKc&e=> > > regards, > Jakob > > > > 2018-04-10 2:50 GMT+02:00 Rahimi, Tariq Rahim < > tariqrahim.rah...@rockets.utoledo.edu>: > > Hello everyone, > > > When I output the tripinfo, it says waitSteps= an integer value. How can I > get this or the total waiting time for each vehicle in seconds? Also, how > can I calculate a queue length? > > > What is the concept of time Step? when I generate a route file with 12 > vehicles departing at the same time step but in different time frame, such > as the first vehicle departs at 0 seconds, the second at 3rd second, the > third one in 6th second but they all are departing at the same time step. > Here, the time step in 9 seconds. 10 minutes later, there are 5 vehicles > departing at the same time step and the time step there becomes 5 seconds. > > > What can I make of time step if I want to calculate the waiting time at an > intersection? > > > Also, any ideas on what does tau trigger? what file can I find it's source > code and what functions are related to it? I want to simulate different > perception-reaction times (0.5 second, 1 second, 2 seconds), How should I > go for it? > > > Thank you, > > > * ------------------------------------- Tariq R Rahimi* > Graduate Student > > > >
_______________________________________________ sumo-user mailing list sumo-user@eclipse.org To change your delivery options, retrieve your password, or unsubscribe from this list, visit https://dev.eclipse.org/mailman/listinfo/sumo-user
