[sumo-user] Measuring Capacity of a Lane

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

[sumo-user] Measuring Capacity of a Lane

Bae, Jong In
Hello,

I ran a simple experiment to measure the capacity of a one-lane signalized exit ramp that is 1500 feet long.
The result came out a lot higher than expected. The capacity came out to be roughly 1,100 veh/hr. (I was expecting to see around 750~800 veh/hr since the ramp is signalized). 
Is this a typical capacity value in SUMO? I've also included a brief scenario settings and methodology below.

The goal was to measure the capacity of a one-lane exit ramp which is 1500 feet long.
A signal (G:50, R:70) is placed at the end of the exit ramp. A single vehicle type was used with minGap=2.2 and tau=1. 
The speed on the exit ramp is 24.59m/s (approximately 55mph). A few iterations were done with various traffic flows.
Each iteration was run for 4500 timesteps where the vehicle count of the last 3600 timesteps was collected (first 900 timesteps were a warm-up period).
e3detectors were placed right after the signal and 'vehicleSum' was used to get the vehicle count for the 3600 timesteps.
The maximum count out of all iterations was determined to be the capacity.

Any help would be greatly appreciated.

Thank you,




_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
Reply | Threaded
Open this post in threaded view
|

Re: [sumo-user] Measuring Capacity of a Lane

Jakob Erdmann
Hello,
my own test shows a flow of ~880.
If the road after the signal continues at 14m/s, flow drops to 780
If there is a turn after the signal, flow drops even further depending on turning radius.
Can you share your scenario?

regards,
Jakob


Am Do., 10. Sept. 2020 um 19:23 Uhr schrieb Bae, Jong In <[hidden email]>:
Hello,

I ran a simple experiment to measure the capacity of a one-lane signalized exit ramp that is 1500 feet long.
The result came out a lot higher than expected. The capacity came out to be roughly 1,100 veh/hr. (I was expecting to see around 750~800 veh/hr since the ramp is signalized). 
Is this a typical capacity value in SUMO? I've also included a brief scenario settings and methodology below.

The goal was to measure the capacity of a one-lane exit ramp which is 1500 feet long.
A signal (G:50, R:70) is placed at the end of the exit ramp. A single vehicle type was used with minGap=2.2 and tau=1. 
The speed on the exit ramp is 24.59m/s (approximately 55mph). A few iterations were done with various traffic flows.
Each iteration was run for 4500 timesteps where the vehicle count of the last 3600 timesteps was collected (first 900 timesteps were a warm-up period).
e3detectors were placed right after the signal and 'vehicleSum' was used to get the vehicle count for the 3600 timesteps.
The maximum count out of all iterations was determined to be the capacity.

Any help would be greatly appreciated.

Thank you,



_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user

_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
Reply | Threaded
Open this post in threaded view
|

Re: [sumo-user] Measuring Capacity of a Lane

Bae, Jong In
Jakob,

Thank you for your reply and taking a look at my scenario.
Please see the attached zip file for the scenario files.
I've also attached the results I got from the e3Detectors where I summed up the 'interval_vehicleSum' (the last 3600s) for the capacity.

Thank you,




From: [hidden email] <[hidden email]> on behalf of Jakob Erdmann <[hidden email]>
Sent: Friday, September 11, 2020 3:34 AM
To: Sumo project User discussions <[hidden email]>
Subject: Re: [sumo-user] Measuring Capacity of a Lane
 
Hello,
my own test shows a flow of ~880.
If the road after the signal continues at 14m/s, flow drops to 780
If there is a turn after the signal, flow drops even further depending on turning radius.
Can you share your scenario?

regards,
Jakob


Am Do., 10. Sept. 2020 um 19:23 Uhr schrieb Bae, Jong In <[hidden email]>:
Hello,

I ran a simple experiment to measure the capacity of a one-lane signalized exit ramp that is 1500 feet long.
The result came out a lot higher than expected. The capacity came out to be roughly 1,100 veh/hr. (I was expecting to see around 750~800 veh/hr since the ramp is signalized). 
Is this a typical capacity value in SUMO? I've also included a brief scenario settings and methodology below.

The goal was to measure the capacity of a one-lane exit ramp which is 1500 feet long.
A signal (G:50, R:70) is placed at the end of the exit ramp. A single vehicle type was used with minGap=2.2 and tau=1. 
The speed on the exit ramp is 24.59m/s (approximately 55mph). A few iterations were done with various traffic flows.
Each iteration was run for 4500 timesteps where the vehicle count of the last 3600 timesteps was collected (first 900 timesteps were a warm-up period).
e3detectors were placed right after the signal and 'vehicleSum' was used to get the vehicle count for the 3600 timesteps.
The maximum count out of all iterations was determined to be the capacity.

Any help would be greatly appreciated.

Thank you,



_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user

_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user

capacity.zip (6K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [sumo-user] Measuring Capacity of a Lane

Jakob Erdmann
Hello,
the difference in observed flows comes from the different step-lengths of the simulation. An explanation is given at https://sumo.dlr.de/docs/Car-Following-Models.html#actionsteplength
Basically, the effect on reaction time affects the flow when starting a queue of stopped vehicles.

regards,
Jakob



Am Fr., 11. Sept. 2020 um 18:14 Uhr schrieb Bae, Jong In <[hidden email]>:
Jakob,

Thank you for your reply and taking a look at my scenario.
Please see the attached zip file for the scenario files.
I've also attached the results I got from the e3Detectors where I summed up the 'interval_vehicleSum' (the last 3600s) for the capacity.

Thank you,




From: [hidden email] <[hidden email]> on behalf of Jakob Erdmann <[hidden email]>
Sent: Friday, September 11, 2020 3:34 AM
To: Sumo project User discussions <[hidden email]>
Subject: Re: [sumo-user] Measuring Capacity of a Lane
 
Hello,
my own test shows a flow of ~880.
If the road after the signal continues at 14m/s, flow drops to 780
If there is a turn after the signal, flow drops even further depending on turning radius.
Can you share your scenario?

regards,
Jakob


Am Do., 10. Sept. 2020 um 19:23 Uhr schrieb Bae, Jong In <[hidden email]>:
Hello,

I ran a simple experiment to measure the capacity of a one-lane signalized exit ramp that is 1500 feet long.
The result came out a lot higher than expected. The capacity came out to be roughly 1,100 veh/hr. (I was expecting to see around 750~800 veh/hr since the ramp is signalized). 
Is this a typical capacity value in SUMO? I've also included a brief scenario settings and methodology below.

The goal was to measure the capacity of a one-lane exit ramp which is 1500 feet long.
A signal (G:50, R:70) is placed at the end of the exit ramp. A single vehicle type was used with minGap=2.2 and tau=1. 
The speed on the exit ramp is 24.59m/s (approximately 55mph). A few iterations were done with various traffic flows.
Each iteration was run for 4500 timesteps where the vehicle count of the last 3600 timesteps was collected (first 900 timesteps were a warm-up period).
e3detectors were placed right after the signal and 'vehicleSum' was used to get the vehicle count for the 3600 timesteps.
The maximum count out of all iterations was determined to be the capacity.

Any help would be greatly appreciated.

Thank you,



_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user

_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
Reply | Threaded
Open this post in threaded view
|

Re: [sumo-user] Measuring Capacity of a Lane

Bae, Jong In
Jakob,

I must be doing something wrong. I changed the simulation step length to 1s and re-ran the experiment.
The maximum hourly vehicle count was still around 1,040 vehicles/hr.
Did you still get the flow ~880 when you ran a test with my scenario?

I wonder if I am using a different methodology or setting than you are.
I think it'd be better to first get the same result as you and then start exploring the effects of using different settings.

Thank you so much for your help!

From: [hidden email] <[hidden email]> on behalf of Jakob Erdmann <[hidden email]>
Sent: Monday, September 14, 2020 10:01 AM
To: Sumo project User discussions <[hidden email]>
Subject: Re: [sumo-user] Measuring Capacity of a Lane
 
Hello,
the difference in observed flows comes from the different step-lengths of the simulation. An explanation is given at https://sumo.dlr.de/docs/Car-Following-Models.html#actionsteplength
Basically, the effect on reaction time affects the flow when starting a queue of stopped vehicles.

regards,
Jakob



Am Fr., 11. Sept. 2020 um 18:14 Uhr schrieb Bae, Jong In <[hidden email]>:
Jakob,

Thank you for your reply and taking a look at my scenario.
Please see the attached zip file for the scenario files.
I've also attached the results I got from the e3Detectors where I summed up the 'interval_vehicleSum' (the last 3600s) for the capacity.

Thank you,




From: [hidden email] <[hidden email]> on behalf of Jakob Erdmann <[hidden email]>
Sent: Friday, September 11, 2020 3:34 AM
To: Sumo project User discussions <[hidden email]>
Subject: Re: [sumo-user] Measuring Capacity of a Lane
 
Hello,
my own test shows a flow of ~880.
If the road after the signal continues at 14m/s, flow drops to 780
If there is a turn after the signal, flow drops even further depending on turning radius.
Can you share your scenario?

regards,
Jakob


Am Do., 10. Sept. 2020 um 19:23 Uhr schrieb Bae, Jong In <[hidden email]>:
Hello,

I ran a simple experiment to measure the capacity of a one-lane signalized exit ramp that is 1500 feet long.
The result came out a lot higher than expected. The capacity came out to be roughly 1,100 veh/hr. (I was expecting to see around 750~800 veh/hr since the ramp is signalized). 
Is this a typical capacity value in SUMO? I've also included a brief scenario settings and methodology below.

The goal was to measure the capacity of a one-lane exit ramp which is 1500 feet long.
A signal (G:50, R:70) is placed at the end of the exit ramp. A single vehicle type was used with minGap=2.2 and tau=1. 
The speed on the exit ramp is 24.59m/s (approximately 55mph). A few iterations were done with various traffic flows.
Each iteration was run for 4500 timesteps where the vehicle count of the last 3600 timesteps was collected (first 900 timesteps were a warm-up period).
e3detectors were placed right after the signal and 'vehicleSum' was used to get the vehicle count for the 3600 timesteps.
The maximum count out of all iterations was determined to be the capacity.

Any help would be greatly appreciated.

Thank you,



_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user

_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user

capacity.zip (11K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [sumo-user] Measuring Capacity of a Lane

Jakob Erdmann
Sorry, I forgot to mention that the value of sigma=0.0 also has a major effect.
The default value of sigma is 0.5 which causes small deviations from the optimal speed. With your sigma value of 0 you get "perfect" car-following which also raises the flow.

Am Mo., 14. Sept. 2020 um 18:54 Uhr schrieb Bae, Jong In <[hidden email]>:
Jakob,

I must be doing something wrong. I changed the simulation step length to 1s and re-ran the experiment.
The maximum hourly vehicle count was still around 1,040 vehicles/hr.
Did you still get the flow ~880 when you ran a test with my scenario?

I wonder if I am using a different methodology or setting than you are.
I think it'd be better to first get the same result as you and then start exploring the effects of using different settings.

Thank you so much for your help!

From: [hidden email] <[hidden email]> on behalf of Jakob Erdmann <[hidden email]>
Sent: Monday, September 14, 2020 10:01 AM
To: Sumo project User discussions <[hidden email]>
Subject: Re: [sumo-user] Measuring Capacity of a Lane
 
Hello,
the difference in observed flows comes from the different step-lengths of the simulation. An explanation is given at https://sumo.dlr.de/docs/Car-Following-Models.html#actionsteplength
Basically, the effect on reaction time affects the flow when starting a queue of stopped vehicles.

regards,
Jakob



Am Fr., 11. Sept. 2020 um 18:14 Uhr schrieb Bae, Jong In <[hidden email]>:
Jakob,

Thank you for your reply and taking a look at my scenario.
Please see the attached zip file for the scenario files.
I've also attached the results I got from the e3Detectors where I summed up the 'interval_vehicleSum' (the last 3600s) for the capacity.

Thank you,




From: [hidden email] <[hidden email]> on behalf of Jakob Erdmann <[hidden email]>
Sent: Friday, September 11, 2020 3:34 AM
To: Sumo project User discussions <[hidden email]>
Subject: Re: [sumo-user] Measuring Capacity of a Lane
 
Hello,
my own test shows a flow of ~880.
If the road after the signal continues at 14m/s, flow drops to 780
If there is a turn after the signal, flow drops even further depending on turning radius.
Can you share your scenario?

regards,
Jakob


Am Do., 10. Sept. 2020 um 19:23 Uhr schrieb Bae, Jong In <[hidden email]>:
Hello,

I ran a simple experiment to measure the capacity of a one-lane signalized exit ramp that is 1500 feet long.
The result came out a lot higher than expected. The capacity came out to be roughly 1,100 veh/hr. (I was expecting to see around 750~800 veh/hr since the ramp is signalized). 
Is this a typical capacity value in SUMO? I've also included a brief scenario settings and methodology below.

The goal was to measure the capacity of a one-lane exit ramp which is 1500 feet long.
A signal (G:50, R:70) is placed at the end of the exit ramp. A single vehicle type was used with minGap=2.2 and tau=1. 
The speed on the exit ramp is 24.59m/s (approximately 55mph). A few iterations were done with various traffic flows.
Each iteration was run for 4500 timesteps where the vehicle count of the last 3600 timesteps was collected (first 900 timesteps were a warm-up period).
e3detectors were placed right after the signal and 'vehicleSum' was used to get the vehicle count for the 3600 timesteps.
The maximum count out of all iterations was determined to be the capacity.

Any help would be greatly appreciated.

Thank you,



_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user

_______________________________________________
sumo-user mailing list
[hidden email]
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user