BISC Seminar Announcement, Tuesday Sept. 22nd, 4-6pm , 310 Soda Hall

Frank Hoffmann (fhoffman@cs.berkeley.edu)
Mon, 21 Sep 1998 23:42:03 +0200 (MET DST)

************************************************************
B I S C S e m i n a r A n n o u n c e m e n t
************************************************************

!!! Special BISC Seminar on Research Activities at the !!!
Institute of Automatic Control,
Darmstadt University of Technology, Germany.
Prof. Rolf Isermann

This special seminar is held in conjunction with
the Mechanical Engineering seminar series and features two talks
from Oliver Nelles and Dominik Fuessel from the Darmstadt University
of Technology preceded by an introduction of Prof. Isermann
who is currently visiting scholar at UC Berkeley.

!!! Notice , due to the schedule of our speakers !!!
this seminar takes place on TUESDAY Sept. 22nd.
There will be no seminar on Thursday Sept. 24th.

*************** First talk ****************************************

Identification of nonlinear processes with dynamic local linear neural
networks (LOLIMOT) and applications for engines and thermal plants

Speaker : Oliver Nelles
Institute of Automatic Control,
Darmstadt University of Technology, Germany
Email: ONelles@irt.tu-darmstadt.de

Date: Tuesday, September 22nd, 1998
Time: 4-5pm
Location : 310 Soda Hall

Abstract

Special dynamic neural networks with external dynamics and local linear
models weighted with normalized radial basis functions will be
presented. This LOLIMOT (local linear model tree) approach can be
trained relatively fast and allows a fuzzy rule interpretation.
Applications will be shown for turbochargers of Diesel engines
and an industrial heat exchanger. LOLIMOT can also be used for the
adaptation of fuzzy-models (grey box models) and for fuzzy model
based control system design, as e.g. fuzzy model predictive control.a

*************** Second talk ****************************************

Fault diagnosis with selflearning fuzzy-neuro systems (SELECT)

Speaker : Dominik Fuessel
Institute of Automatic Control,
Darmstadt University of Technology, Germany
Email: dfuessel@irt.tu-darmstadt.de

Date: Tuesday, September 22nd, 1998
Time: 5-6pm
Location : 310 Soda Hall

Abstract

Methods of model based fault detection generate deviations from
normal behaviour, called symptoms. These symptoms can be generated
via methods of parameter estimation, neural networks or parity
equations. They are the basis of fault diagnosis. However, the
fault-symptom trees required for reasoning are difficult to obtain.
The presentation shows how causal fault-symptom trees can be learned
by experiments with real systems. A new procedure which combines the
learning ability of neural networks with the transparency of fuzzy
logic systems is a SElf LEarning Classification Tree (SELECT).
By measurements it creates neural decision nodes. The resulting
tree can be augmented with a priori knowledge and parameter-opti-
mization allows fine-tuning. Applications are shown for fault-
diagnosis of a d.c.motor with circulation pump and a thermal plant.

**********************************************************************
Please direct questions with regard to the contents of the talk
and request for papers to the speaker. To unsubscribe send email
to fhoffman@cs.berkeley.edu NOT !! to bisc-group@diva.eecs.berkeley.edu
**********************************************************************

-- 
---------------------------------------------------------------------------
Frank Hoffmann                               UC Berkeley
Computer Science Division                    Department of EECS
Email: fhoffman@cs.berkeley.edu              phone: 1-510-642-8282
URL: http://http.cs.berkeley.edu/~fhoffman   fax:  1-510-642-5775
---------------------------------------------------------------------------

############################################################################ This message was posted through the fuzzy mailing list. (1) To subscribe to this mailing list, send a message body of "SUB FUZZY-MAIL myFirstName mySurname" to listproc@dbai.tuwien.ac.at (2) To unsubscribe from this mailing list, send a message body of "UNSUB FUZZY-MAIL" or "UNSUB FUZZY-MAIL yoursubscription@email.address.com" to listproc@dbai.tuwien.ac.at (3) To reach the human who maintains the list, send mail to fuzzy-owner@dbai.tuwien.ac.at (4) WWW access and other information on Fuzzy Sets and Logic see http://www.dbai.tuwien.ac.at/ftp/mlowner/fuzzy-mail.info (5) WWW archive: http://www.dbai.tuwien.ac.at/marchives/fuzzy-mail/index.html