A fuzzy cerebellar model articulation controller

Abstract

An adaptive fuzzy controller based on the cerebellar model articulation controller is presented. The controller is basically a table lookup module in which fuzzy sets are stored and manipulated by using fuzzy set theory. Analogous to neural networks, this controller adaptively estimates continuous functions from sampled data without specifying mathematically how outputs depend on inputs. As a result, it is easy to modify in order to accommodate changes in the system. Learning is based on observations of the input-output relationship of the system, and there is no need to identify fuzzy logical rules from the human operator. The goal is to demonstrate the use of a neural-network scheme as applied to building the adaptive fuzzy system.<>