A fuzzy learning system emulating a human tracking operator

The authors describe a recursive fuzzy estimator in fuzzy state machine terms and a probabilistic fuzzy identifier, which together form elements of a fuzzy modelling technique capable of emulating a prototype dynamic system, of which only the input-output behaviour is known. The model was successfully applied to emulate the behaviour of a human controller in a tracking loop. It is indicated how this learning system can be implemented in terms of a number of computational formations organized in a parallel processing architecture