Tuning a Fuzzy Logic Controller: On the GARIC Architecture
F.P.G. Segers · TU/e Research Portal · 1994
In this paper the GARIC (Generalized Approximate Reasoning-based Intelligent Control) architecture, as proposed by Berenji and Khedkar [5 ] , wil l be investigated. GARIC learns to control a system by means of reinforcement learning. The fuzzy controller in GARIC learns by tuning its membership functions. Attention will be paid to the operatio-. sf GARIC =BI ts hey: ai.V what it !earns. The GAXIC architectwe is applied to a rotationJtranslation-robot to illustrate its operation.