Analog processing for nonlinear fuzzy controller development

Abstract

In this work we present some analog circuits used in fuzzy control. We have done some modifications to conventional implementations and we present a new circuit for the inference stage. We have substituted the conventional method of rule aggregation for a simpler one, which results in a reduced number of transistors. We also present an example of an analog fuzzy controller applied to a nonlinear plant with multiplicity of stable and unstable equilibrium points, different operation regions and different operation modes.