From: earlcox@earlcoxreports.com   
      
   Bill,   
      
     If by "All of these comments" you mean mine, I am unsure of your point. My   
   examples had nothing to do with control, in fact they were addressed toward   
   the kinds of fuzzy logic applications that I have been building and   
   delivering to paying customers since the very early 1980's. These include   
   models in portfolio safety and suitability, adaptive pricing, project risk   
   assessment, infrastructure performance optimization, multi-objective and   
   multi-constraint scheduling, network topology configuration, criminal   
   (suspect) identification, global container routing and transshipment, and   
   many, many more. If by "All of these comments" you mean the general threads   
   that appear on this news group -- I totally agree!   
      
     Control engineering applications have their place. I was one of the prime   
   developers of Motorola's fuzzy logic tutorial for embedded (control)   
   applications. I have seen very powerful fuzzy control systems that worked as   
   well and often better than their PID counterparts.   
      
      However, that said, the problem in this news group, Bill, is NOT that it   
   is dominated by control engineering types. The problem is that is it   
   dominated by academics discussing fuzzy set theory and fuzzy (approximate)   
   reasoning. I would welcome discussions on fuzzy control engineering -- how   
   to build robust FAMs, how to generate second order rules, the best   
   techniques for aggregation, the methods for defuzzification, how to code   
   8-bit HC11 engines to fit in some small physical plant, etc. etc. But if you   
   go back over the past three years of this news group, I defy you to find ANY   
   discussion of the practical applications of fuzzy logic in control (other   
   than the ubiquitous queries from grad students on Wednesday  trying to   
   finish (or start) a project that is due on Friday).   
      
      Patrick's question about automatic gain control at least opens the door   
   to a real world problem. In such a context many interesting and subtle   
   concepts in fuzzy logic an be illuminated and explored.   
      
      IMHO,   
      E a r l   
      
      
      
   "Bill Silvert"  wrote in message   
   news:1118166015.435226.194750@f14g2000cwb.googlegroups.com...   
   > All of these comments assume that if fuzzy logic is not used in an   
   > engineering application it is too wishy-washy to be meaningful. I   
   > disagree. I've been using fuzzy logic to assess environmental impacts   
   > and to evaluate qualitative data like diver surveys of the seabed, as   
   > well as scoring forecasting skills.   
   >   
   > It doesn't have to be on a chip to be useful. But I think that this   
   > list attracts pretty hard-core engineering types and I haven't seen   
   > much in the way of tangential applications. (And then there are the   
   > spammers.)   
   >   
   > I guess the situation is similar to what one would find with a   
   > statistics list - there would be a lot of statisticians, but not too   
   > many people who use statistics in other disciplines.   
   >   
   > Bill Silvert   
   >   
   > EarlCox wrote:   
   > > Fuzzy logic has   
   > > wide applicability in control theory applications including AGC...   
   >   
   > > "patrika"  wrote in message   
   > > > I have been thinking that fuzzy logic might be good to use on an   
   > > > embedded system that uses automatic gain control (AGC). >   
   >   
   > > EarlCox wrote:   
   > > no one is discussing problems with the actual   
   > > use   
   > > > > of fuzzy logic in real world applications.   
   >   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)