Evacuation time analysis

All topics related to questions in Pedestrian Dynamics
ie_1989
Posts: 58
Joined: Saturday 21 November, 2015 - 06:50

Re: Evacuation time analysis

Post by ie_1989 »

Hi. I have run into a problem in evacuation time. In my model an Action Timer starts at 6:00 minutes of the simulation. But when it starts some agents got stuck at escalators. I have used an If statement for agent route selection when the evacuation starts. Does this statement produce the problem? I have attached my model.
Thank you in advance
Attachments
new_project-2.rar
(268.42 KiB) Downloaded 131 times
Bram de Vries
Posts: 60
Joined: Thursday 08 January, 2015 - 13:29

Re: Evacuation time analysis

Post by Bram de Vries »

Dear ie_1989,

Your problem is caused by the fact that you are running with the setting Re-route periodically (in the Agent profiles). This is unfortunately not possible in combination with the Action Timer which alters the directions of your escalators and that's why some Agents get stuck on them.

I would advise you not to use this setting. If you want an Agent to re-plan his route at a certain point, you could use an Action Area at that location with a trigger with the function Agent_StartRouting(i)

Kind regards,

Bram
ie_1989
Posts: 58
Joined: Saturday 21 November, 2015 - 06:50

Re: Evacuation time analysis

Post by ie_1989 »

Thank you very much for your response. I set all escalators Bidirectional when Action Timer starts and it got solved. I need agents reroute periodically because otherwise they will accumulate near some specific passage ways (which I can't figure out why).
Bram de Vries
Posts: 60
Joined: Thursday 08 January, 2015 - 13:29

Re: Evacuation time analysis

Post by Bram de Vries »

Dear ie_1989,

I'm happy to hear that your problem is solved! If you let me know where these accumulations take place I could take a look at it.

Kind regards,

Bram
ie_1989
Posts: 58
Joined: Saturday 21 November, 2015 - 06:50

Re: Evacuation time analysis

Post by ie_1989 »

Dear Bram
In Layer ID 3 (Named 'E3 Ticket') I have located numerous passageways which are one-directional before the action timer starts. After starting the evacuation, they all become bi-directional. But the agents choose some specific ones (mainly the ones close to the blocks) to exit and they accumulate around them while there are plenty of unused passageways.
I have another question about way points logs. I have attached my new model here. In my model, at time 7:15 the evacuation starts. Meanwhile, 1000 number of agents are generated at this time (the red ones) who should be evacuated. I have used 3 action timer 2 of which are just for making some passageways one-direction. The main one named "Action Timer_1" is the one in which I collect the agents' log by means of a way point. I also defined a new route in the output graphs which collects the data on the red agents only with the profile id of "Emergency Incoming". The statistics gathered from the way point is different than the ones from this in their exit time, while I expect the max time of exit should be the same for both. Why does this happen?
ie_1989
Posts: 58
Joined: Saturday 21 November, 2015 - 06:50

Re: Evacuation time analysis

Post by ie_1989 »

I have attached my file here
Attachments
Myversion - Final.rar
(306.23 KiB) Downloaded 125 times
Bram de Vries
Posts: 60
Joined: Thursday 08 January, 2015 - 13:29

Re: Evacuation time analysis

Post by Bram de Vries »

Dear ie_1989,

With regards to the accumulation at the passageways. I would suggest to also use an Action Area with Agent_StartRouting(i) over there. This will have the same effect as using Reroute periodically. Especially since you have set the update time at 10 seconds. If you locate the action area close enough, the agents will reach the passageways within 10 seconds.

As for the loggings, you have selected in ActionTimer_1 in the Effect on agents tab that the trigger should be executed for all Agents (Your Condition is TRUE). I would change this to Agent_GetProfileID(s) = 7, because then only these 1000 agents are triggered. There is one more thing you should do and that is to create these 1000 agents not at time mins(7.25) but at mins(7.249) for instance, because the trigger to log them is now executed before they are created.

Kind regards,

Bram
ie_1989
Posts: 58
Joined: Saturday 21 November, 2015 - 06:50

Re: Evacuation time analysis

Post by ie_1989 »

Dear Bram
Thank you very much for your help. I used action areas but it does not work as good as the state in which the reroute planning is enabled. But the solution for way point logs. Thanks a lot
ie_1989
Posts: 58
Joined: Saturday 21 November, 2015 - 06:50

Re: Evacuation time analysis

Post by ie_1989 »

another question. Is it possible to remove agents in specific layers when the action timer starts?
Bram de Vries
Posts: 60
Joined: Thursday 08 January, 2015 - 13:29

Re: Evacuation time analysis

Post by Bram de Vries »

Dear ie_1989,

Could you maybe clarify what you mean by removing the agents? Do you want them to be 'destroyed' as soon as the action timer goes off?

Kind regards,

Bram
Post Reply