Randy wrote

> The whole point of Fuzzy Logic is to erase the possibility of overshooting, or
 at least to minimize
> it. With fuzzy, you use "membership functions" to determine the "degree of mem
bership" (how "true" a
> condition is - from 0 to 1) in a set (such as "Too Hot", "Getting Warmer", etc
.). The boundaries for
> these sets are not deninite, but variable - a value can have a partial members
hip in 2 sets. By
> tweaking with values for your output and input membership functions, you can g
et an accurate
> controller without much (or even any) mathematical modelling.
>
Agreed... its great for a control scheme like this.

Since I started the ball rolling on fuzzy or other control schemes..

What I originally was trying to point out was that fuzzy logic is hard
to get youre neurons around.. as in from theory to actual code
without some sort of development system such as Microchips
fuzzyLAB. I could not get started with it until I purchased fuzzyLAB
anyway... I guess I am probably a bit thick.
So if you want to spend the money.. go for it. Its really a good
development environment.

I was then trying to say that for a simple temperature control
scheme like that presented you could probably get some good
results from using lookup tables. Maths could be avoided in favour
of experimentation. Although I suspect the best control curve will
look like a half sine wave, the peak being the normalized desired
temperature setting.



_____________________________

Lance Allen
Technical Officer
Uni of Auckland
Psych Dept
New Zealand

www.psych.auckland.ac.nz

_____________________________