Thunderhead Engineering Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Forum moved to https://forum.thunderheadeng.com

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - SFC

Pages: [1]
1
Pathfinder / Future Release Wishlist
« on: October 07, 2021, 12:13:35 am »
Hi,

I have been using Pathfinder for a while and think it is a very impressive piece of software. There are some features I'd like to see in future and thought that others could chime in with theirs, too. I did find an older post about a wishlist but it didn't seem to catch on. There was a link to some other forum to talk about future features but this link was dead. Is there somewhere else we should discuss this?

In the meantime, here goes:

I'd like to see some sort of devices class that could be used to control behaviour switching and door availability. For example, instead of just having a door availability switch at a predetermined time as inputed to the table in the door properties, the door could be tied to a central timer and the availability changes are triggered based on that. That way if you had multiple doors that are to follow the same schedule but you want to change the schedule you could just edit the timer rather than find all the doors you want to affect and change their schedules. Maybe the trigger for these could be more complicated than just a timer, i.e. a door availability based on the number of occupants still in the room, or the flow rate through the door. I guess I'm thinking of something like the control logic in FDS.

How about scheduled (or triggered as above) accepted profile properties for doors/rooms? Just another level of evolving complex behaviour.

An idle wandering function, or random walk behaviour where an occupant enters a space and waits, but rather than following the current wait protocol of standing still and moving away from active doors etc, they just spread out randomly and walk slowly.

I have been using the new feature of attractors as triggers for behaviour changes based on occupants entering a room and these have resulted in some resonably complex behaviour, at least complex enough for what I wanted. So this is a nice feature and it has given be the taste for more complex modelling.

Please take this as an appreciative wish list, not a list of demand. Let me know if there are ways I have overlooked to achieve the desired functionality within the current model.

Cheers.






2
I'll note that the original problem I had still exists: occupants positions being randomised throughout the whole model as opposed to their rooms of origin.

Is this forum still monitored by the developers?

3
Hi,

I see the latest release of Pathfinder has the the following release note:

When using the command line monte carlo simulation generator, additional parameters now provide control over property and position randomization.

This seems to be exactly what I wanted re my previous post. Is there any documentation for usage? The bat file contains the lines:

:: Additional filtering options are available with the format
:: -include=Foo\Bar -exclude_pos=Baz

...but I have no idea what the correct option syntax to use would be. Any help?

4
Hi,

I wanted to use the montecarlo_createcases.bat to analyse the effect of occupants initial starting positions on the floor and overall evacuation times. I have noticed that changing the position of a single occupant in a room with 200 occupants can change the local and global evacuation times up to 10s of seconds. This makes it difficult to select simulation runs for comparison. Is it the worst case? How can I tell? The effect is also greater than I would have expected.

I hoped that Monte Carlo would help with this problem and I was pleased to see the facility available (although I couldn't find any mention of it in any of the documentation provided with Pathfinder). The problem I am having now is, while the occupants positions are being randomised, they are being randomised around the entire model, rather than randomised in their room of origin. This is unfortunate as I need to have each floor level's occupants assigned to a profile that is specific to that floor. This is in order to restrict which egress elements they use when they leave their original floor, by assigning "accepted profiles" to doors on other floors, that will not admt them. 

I can't see anyway I can affect what parameters are randomised, and how they are randomised. Is there a way?

As a side, while at some stage in the future I may be interested in modelling populations with distributed parameters, at the moment I just want to have all occupants with the same movement parameters. I'm pretty sure this is the case, but can you please confirm that if I use the montecarlo batch file, it won't modify parameters that are set to constant?

Thanks in advance.


Pages: [1]