home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!sun-barr!ames!agate!dog.ee.lbl.gov!epb2!envwpk
- From: envwpk@epb2.lbl.gov (Werner P. Keilholz)
- Newsgroups: comp.ai
- Subject: Intelligent Simulation Environments - request
- Date: 10 Sep 1992 11:55:48 GMT
- Organization: Lawrence Berkeley Laboratory, Berkeley
- Lines: 65
- Distribution: world
- Message-ID: <26201@dog.ee.lbl.gov>
- Reply-To: envwpk@epb2.lbl.gov (Werner P. Keilholz)
- NNTP-Posting-Host: 128.3.12.39
-
-
- I have just started a PhD work on 'Intelligent Simulation Environments'.
- The purpose of this work is to simplify the use of building performance
- evaluation codes (simulators) such as TRNSYS, COMIS or others (in particular,
- general purpose solvers such as SPANK/SPARK). To achieve this goal, a generic
- graphical environment will be developped. A prototype of such an environment
- already exists for TRNSYS.
-
- To bring a certain degree of intelligence into such a system, an expert
- system layer will assist the user in describing his simulation problem
- to the system and help the user to choose the simulation program(s) best
- adapted to this problem.
- Another component of the system will allow future users to couple the
- simulation tools made available through the ISE, i.e. run the same project
- (e.g. the simulation of an existing building) with different simulation codes,
- so that the user has to enter data shared by different programs only once.
- Furthermore coupled simulation runs should be possible using the ping-pong
- coupling method.
-
- In this context, I would be glad to get *any* information about ongoing
- or finished projects, articles, books, etc. that might be of interest for
- us. In particular, we are looking for:
-
- * expert knowledge (rules, rule bases, examples) about how to assemble TRNSYS
- models to solve a given problem (i.e., to reach a given simulation goal)
- * expert rules for making automatic connections of TRNSYS models (once the
- models are connected, the *variables* of the models have to be connected
- as well; we can imagine that this task could be automized to a certain
- degree)
- * All of the above for other simulation codes than TRNSYS (COMIS, HVACSIM, ...)
- and general rules applicable to all simulators
-
- * experience in coupling simulation codes and expert rules about
- questions like, 'in what circumstances is ping-pong- coupling sufficient ?'
- 'When do convergence problems typically occur ?'
-
- * Ideas about what you, as a possible user of such an environment, would
- like to see implemented in such a system
-
- I'd be happy to get any feedback. In particular, if you are interested in
- details about the project, I'd be glad to send you more detailed
- information.
-
- Cooperation with building professionals using TRNSYS or COMIS
- would be extremely useful and of benefit for both parties involved.
-
-
- Thanks for your time,
-
- Werner
-
-
- P.S.: Please reply via email to envwpk@epb2.lbl.gov, since I do not have !
- access to the INTERNET while in France. Do not reply to this newsgroup, !
- I propably won't be able to read it any more. !
-
- -------------------------------------------------------------------------------
- Werner Keilholz | Email: envwpk@epb2.lbl.gov | \ | /
- CSTB | or : werner@cstb.fr | \ /\| /
- B.P. 209 | | \/ \/
- 06904 Sophia Antipolis Cedex | Phone: (FRANCE) 93653400 | |\
- FRANCE * United States of Europe | Fax: (FRANCE) 93652937 | | \
- -------------------------------------------------------------------------------
-
-
-