ABSTRACT
Recently many new products equipped wi th intelligent controllers have been launched on the market, a lot o f effort has been made in R & D departments around the world, and numerous papers have been written on how to apply neural networks (NN), fuzzy logic models ( F L M ) , and related ideas o f learning (in classic control terms o f adaptation) for solving control problems o f both linear and non linear systems. N N based algorithms have been recognized as attractive alterna tives to the standard and wel l established adaptive control schemes. Due to their wel l known ability to be a universal approximator o f multivariate functions, N N and F L M are o f particular interest for controlling highly nonlinear, partially known and complex plants or processes. Many good and promising results have been reported and the whole field is developing rapidly but it is still in the initial phase. This is mainly due to a lack o f firm theory despite the very good and seemingly far reaching and important experimental results.
