Re: Fuzzy vs Neural


Subject: Re: Fuzzy vs Neural
pandeyashish@my-deja.com
Date: Fri Jan 07 2000 - 16:04:38 MET


Generally processes are cool because of large time-constents. First
thing you must know before going into fuzzy control is:

There exists no proof of stability. It means that you have no proof
what so ever that the fuzzy logic controller in closed loop is going to
remain stable or not.

Since model based controls for a process like the one you described are
well established I see no reason that you would like to venture into a
grey area.

I think you must have a good reason for using a FLC( this also remains
a grey area) Dont implement FLC just cause they sound cute as there is
no justification for their use in processes where mathematical models
exists.

This is my opinion, I can elaborate but I fear offending everyone in
the group.

Bye
Ashish
achich_p@hotmail.com

In article <386ABDCE.69595C4B@earthlink.net>,
  "Lynn B. Hales" <lhales@earthlink.net> wrote:
> Will, we do real-time process control for a living. Check out our
> www.kscape.com.
> Our product is called KnowledgeScape and it is a real-time expert
control
> system that has both crisp and fuzzy rules, neural networks and
genetic
> algorithms all integrated together into a great system. Lynn Hales
>
> Will Dwinnell wrote:
>
> > Lilzeke wrote:
> > "I work for a major beverage company, and I just got assigned to
figure
> > out the best way to control a vessel(mainly temp and pressure). We
are
> > trying to figure out the best way to control the vessel with new
> > technology in the advanced controls arena. Can anyone recommend a
> > book(s) or website(s) that would get me started in the right
direction
> > with respect to Fuzzy Logic and Neural Neworks.
> > In my brief research, Fuzzy Logic is a great decision maker, but a
> > Neural Network can tightly control a system by supervised or
> > unsupervised 'learning'. Then again, I could be utterly wrong."
> >
> > Both fuzzy logic and neural networks can be applied many different
> > ways. As used in practice, fuzzy logic generally involves
translating
> > some human-level description of the solution to fuzzy code whereas
> > neural networks learn to solve the problem from historical examples.
> >
> > For a really good introduction to rule-based fuzzy logic, see "The
Fuzzy
> > Systems Handbook" by Earl Cox, published by AP. An excellent
> > introduction to machine learning methods (including some material on
> > neural networks) is "Computer Systems That Learn", by Weiss and
> > Kulikowski, published by Morgan Kaufmann.
> >
> > Will Dwinnell
> > predictor@compuserve.com
>
>

Sent via Deja.com http://www.deja.com/
Before you buy.

############################################################################
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



This archive was generated by hypermail 2b25 : Thu Apr 06 2000 - 15:59:36 MET DST