Actually, the number of possible rules grows exponentially with the number of
inputs, i.e. n inputs with m fuzzy sets -> m^n rules.
What one could do about this "curse of dimensionality" [Bellman] (and what I do
:) ) is to use n-1 2-dimensional rulebases instead of one n-dimensional and
arrange them as a binary tree. Some benchmarking with this stuff is already
done and it works quite well. If you are interested, you can download two
papers about it at
http://www.forwiss.uni-erlangen.de/~rrholve/papers.html
Regards
Rainer