Dear Chuck:
I would be happy to try to help out, but would probably need
to see the model to understand what it does (or doesn't do).
I suspect that you need a purely DE model, rather than a
hybrid CT/DE model, which might simplify things.
As for parameter scoping: Every parameter is visible
to everything below it in the hierarchy. It is not
advisable, however, to change values of a parameter
above you in the hierarchy (it may be possible, but
I would have to look up the syntax, and would prefer
not to, since it will cause trouble later). The
problem is that you may get unexpected nondeterminism
(where the result of an execution depends on arbitrary
decisions that the scheduler makes). Changing parameter
values below the ModalModel is fine... The semantics
here is clean and well-defined. You can do this
using transition actions on the ModalModel transitions.
If the parameter named y is inside a mode named x
then the transition action reads "y.x = new value".
Hope this helps. Feel free to send me (or the list)
a model to look at.
Edward
At 01:22 PM 4/20/2006, Lutz, Charles D wrote:
Hi Ptolemy hackers,
Heres the first post, likely of many to come.
We are new Ptolemy users here, although
personally Ive followed the project for several
years (but never got around to diving in!).
Were trying to figure out a way to use Ptolemy
as a general purpose M&S tool, and a
modeling strategy for things like usage of
resources. A scenario could be something
like the self-checkout systems at supermarkets
typically there are 4 or so stations
available. When all are in use, customers wait
in line until one becomes available.
We might model each customers behavior with a
state machine, but then we need a
global variable for the number of available
stations, which would be incremented and
decremented by the (concurrent) instances of
customers. A (waiting) customer can
then transition to e.g. a checkout state when
this variable > 0 and decrement it on
the way.
Ive always had the impression that Ptolemy has
a lot of depth, and now, getting into the
details for real, that is an understatement! We
are trying to get up to speed on how things
work, so as we dont know the theory yet, there
could be problems with this approach, but
we want to try something out, even if it just
helps us understand the dynamics.
We have a simple model with which were trying to get a grasp on some basics.
At the top of the model, we have a
MultipleInstanceComposite with a MIC_out port to a
Display. Also, there is a stationsAvailable parameter, assigned value 4.
Inside the MIC is a ContinuousClock connected to
the trigger port of a ModalModel;
the MM has output ports stations and bagging, going to MIC_out.
Currently the only output actions are to write
some test strings to the MMs output ports.
Issues were having:
* Cant seem to figure out how to get the
concurrent FSMs to be able to see the top
* level parameter (again, could be a bad idea, just trying it for now
)
* Using DEDirector, I got a complaint that
something with a HSDirector can only
* be used in a CTDomain; dont know what
HSDirector is, but I switched to CT
* No output on Display
We are continuing in the manuals, but wanted
perhaps to get a bit of a reality check
Thanks in advance for any input / tips / warnings / etc
Chuck Lutz
Lockheed Martin
Systems of Systems - Modeling and Operations Analysis
BMC4I Modeling and Simulation
Moorestown, NJ
(856)638-7234 (office)
<mailto:[EMAIL PROTECTED]>[EMAIL PROTECTED]
Everyone spoke of an information overload, but what there
was in fact was a non-information overload.
Richard Saul Wurman
------------
Edward A. Lee
Professor, Chair of the EE Division, Associate Chair of EECS
231 Cory Hall, UC Berkeley, Berkeley, CA 94720-1770
phone: 510-642-0253 or 510-642-0455, fax: 510-642-2845
[EMAIL PROTECTED], http://ptolemy.eecs.berkeley.edu/~eal
----------------------------------------------------------------------------
Posted to the ptolemy-hackers mailing list. Please send administrative
mail for this list to: [EMAIL PROTECTED]
