Discrete Compartmental System

This forum contains all archives from the SD Mailing list (go to http://www.systemdynamics.org/forum/ for more information). This is here as a read-only resource, please post any SD related questions to the SD Discussion forum.
Locked
"bjobj"
Junior Member
Posts: 4
Joined: Fri Mar 29, 2002 3:39 am

Discrete Compartmental System

Post by "bjobj" »

Hi

I should like to thank those that have reponded to my questions about developing array models of compartmental systems. It is obvious from their letters that it is easy to model such systems in Vensim, Stella Research and ithink, as well as in Powersims Constructor. It does not seem to be as easy to design an array model of a compartmental system if it is to be possible to assign minimum and/or maximum values to the content of the compartments. I am hard at work trying to integrate the material contributed by the various correspondent in their letters or in the models made available to the members of the mailing list.

If a compartmental system contains discrete entities like people or patients, then, as long as the number of these entities in each compartment is relatively large, the system can be approximated by a continuous model. In some of the systems that I am interested in modeling, the number of entities in some of the compartments is rather small. Such systems should be modeled by discrete models. I should therefore like to pose the following questions:

How well do system dynamics programs like those mentioned above lend themselves to discrete modeling?

What would a discrete array model of a compartmental system look like?

Would it be possible to assign minimum and/or maximum values to the content of the compartments in such a model?


Regards,

Oddur
From: "bjobj" <bjobj@online.no>
"HARRIS,BILL (HP-LakeStevens,ex1
Junior Member
Posts: 5
Joined: Fri Mar 29, 2002 3:39 am

Discrete Compartmental System

Post by "HARRIS,BILL (HP-LakeStevens,ex1 »

I dont want to contradict the idea raised earlier that successful modeling
is focused on solving a problem, not modeling a system.

Having said that (the obligatory disclaimer), you might want to explore
Colored Petri Nets. (No, this is not meant to restart the debate on
discrete event simulation vs. SD! :-) .) If you put feedback and delays
into the system, you have a discrete "SD" model. (Of course, you can do
discrete simulation with many of the SD packages, too.) DesignCPN seems to
be one good tool that should have enough flexibility to handle the arrays
and the constraints (min and max).

Of course, there is something to be said for the clarity of SD models (or
CPN models, for that matter) in which the model layout parallels the system
being modelled.

Bill
--
Bill Harris Year 2000 Program Office
mailto: bill_harris@am.exch.hp.com Hewlett-Packard Company
phone: (425) 335-2200 M/S 330C
fax: (425) 335-2483 8600 Soper Hill Road
home mailto: harris6@halcyon.com Everett, WA 98205-1298
Locked