[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [EnergyPlus_Support] Re: GenOpt and the group simulation
I have found that really using optimization with EnergyPlus requires a
whole lot of effort on preprocessing input files and postprocessing
results.
On the input side, if you want to build up your idf files using macros,
then GenOpt would operate on the imf driver file that has the macro
calls (or possibly an .imf include file). If you have some other way of
creating idf files, then GenOpt would need to operate on a text file
that can control creation of the idf files. If you have simple
perturbations then GenOpt may be able to operate on the *.idf file
directly.
On the results side, remember that GenOpt is "general." It doesn't
require the result of the objective function be in an *.eso, or even any
output file created by EnergyPlus. GenOpt can work with the objective
function you defined below but you have to get the programs to produce
the result. You will need to develop a method of post processing
results using one or more external programs and have GenOpt operate from
the new file created by post processing. You have formulated a pretty
complicated objective function (especially combining results from
multiple Energyplus runs) and I think you will end up having to create
your own post processing program. There are many different programming
languages you could use (perl, python, matlab etc. ) but you'll need to
be able to execute the post processor from a command prompt and without
any GUI interactions.
________________________________
From: EnergyPlus_Support@xxxxxxxxxxxxxxx
[mailto:EnergyPlus_Support@xxxxxxxxxxxxxxx] On Behalf Of natasamdj
Sent: Wednesday, October 11, 2006 1:25 AM
To: EnergyPlus_Support@xxxxxxxxxxxxxxx
Subject: [EnergyPlus_Support] Re: GenOpt and the group simulation
Dear,
Last evening I tested your example with the group simulations. I
found some error, but ok. I saw that there is not a common file as
output; instead there is the same number of the output files as the
number of the input files. I think that GenOpt requires a unique
*.eso file as output to be able to get the necessary data. My
wondering about this group simulation is if it is possible to get
the unique output file form the group simulation.
My second question was about re-running the daily simulation. I
agree with everything that you wrote to me about daily simulation
and Daily frequency on report variables for the run period. But my
aim with all this is bit different. I have the historical energy
consumption data with the daily time step. Exact mathematical way is
adjust the model with the real data is to make optimization for some
model important variables in each point so that the model is as
close as possible to the real data. It is possible to do in MatLab
by using some of functions for non-linear quadratic optimization,
when the objective is sum(Qmi - Qri)* (Qmi - Qri), where i is time
sample, and m is model result and r is real data. In GenOpt it is
not possible to define such an optimization. Therefore I am trying
to make something similar. If GenOpt would optimize some parameters
in each day (that is time sample in my case) so that the sum of
difference in run period is minimal, then the model for each day in
run period is necessary. If I understand using of GenOpt well until
now, that is how such a problem could be solved. If you have some
better idea you can suggest.
Thank you in advance!
Natasa M.Djuic
--- In EnergyPlus_Support@xxxxxxxxxxxxxxx
<mailto:EnergyPlus_Support%40yahoogroups.com> , "Griffith, Brent"
<brent_griffith@...> wrote:
>
> I am not sure I follow all your questions very well, but here are
some
> general responses:
>
>
>
> 1. If you are using GenOpt, then GenOpt would handle and run all
> the parametric simulations and you wouldn't need to use EP-launch
at
> all. You would probably want to set up GenOpt to launch a separate
> batch file that will call EpMacro directly to process *imf files
before
> launching EnergyPlus.
> 2. why rerun consecutive daily simulations at all? If it is for
> run time, note that EnergyPlus has to first do a number of warm up
days
> to setup the thermal histories in the building fabric. You are
better
> off running consecutive days in a single simulation. If it is for
> reporting, then you could choose Daily frequency on report
variables and
> get individual values for each day that way.
> The point of using *.imf is build up *.idf files. So, no,
you
> wouldn't already have a set of *.idf files. They are created by
> successive calls to EpMacro to process the *.imf files.
>
>
>
> Good luck!
>
>
>
> brent
>
>
>
> ________________________________
>
> From: EnergyPlus_Support@xxxxxxxxxxxxxxx
<mailto:EnergyPlus_Support%40yahoogroups.com>
> [mailto:EnergyPlus_Support@xxxxxxxxxxxxxxx
<mailto:EnergyPlus_Support%40yahoogroups.com> ] On Behalf Of natasamdj
> Sent: Tuesday, October 10, 2006 5:33 AM
> To: EnergyPlus_Support@xxxxxxxxxxxxxxx
<mailto:EnergyPlus_Support%40yahoogroups.com>
> Subject: [EnergyPlus_Support] GenOpt and the group simulation
>
>
>
> Dear,
> I would like to ask you some questions about the optimization of
the
> group simulations by using GenOpt.
> 1. Can a series of parametric runs, obtained by using the group
> simulation feature in EP-Launch, be optimized by a couple of a
> command and an initial file from GenOpt? In addition, is it
possible
> to make only an output file from the series of parametric run by
> using postprocessing? Can the result from such a common output
file
> be used as the objective function in GenOpt?
> 2. This question is about re-running daily simulation. If the
group
> of simulations would be for every day in a certain period (from 1
to
> 10 January, for example January 1, January 2, etc), should the
*.idf
> files be needed for each of these days? Is it possible to make an
> *.idf file for every day by using Input Macros? I downloaded your
> example Parametric-5ZoneWarmest.imf. There I found that you have
> made the simulations for the different buildings orientations. I
> wonder if you already had these *.idf files for each of the
> orientations, or *.imf would make them. (Sorry, in the moment I am
> running some optimizations, and I am not capable to run your
example
> now, maybe then I will see well). In my case the parameter will be
> the run period, I assume. I wonder if it is possible to make an
> external file for the different days to be the run period. Then I
> would need the repeating for each day. In such a case can I use a
> recursion:
> "file1.inp contains ##include file2.inp
> file2.inp contains ##include file1.inp"
> as you give the explanation for Input Macros at the page 153 in
> Auxiliary EnergyPlus Programs. Sorry, do I think in good way about
> this option for Macro Input?
>
> If this last about re-running is not clear enough I can write that
> again.
>
> Thank you in advance!
>
> Natasa M.Djuric
>
>
>
>
>
> [Non-text portions of this message have been removed]
>
[Non-text portions of this message have been removed]
The primary EnergyPlus web site is found at:
http://www.energyplus.gov
The group web site is:
http://groups.yahoo.com/group/EnergyPlus_Support/
Attachments are not allowed -- please post any files to the appropriate folder in the Files area of the Support Web Site.
EnergyPlus Documentation is searchable. Open EPlusMainMenu.idf and press the "search" button.
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/EnergyPlus_Support/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/EnergyPlus_Support/join
(Yahoo! ID required)
<*> To change settings via email:
mailto:EnergyPlus_Support-digest@xxxxxxxxxxxxxxx
mailto:EnergyPlus_Support-fullfeatured@xxxxxxxxxxxxxxx
<*> To unsubscribe from this group, send an email to:
EnergyPlus_Support-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/