Position Control of Servo Hydraulic Cylinder Using a Neuro-Fuzzy Controller with On-Line Searched Parameters

Abstract

The purpose of this study was to design a neuro-fuzzy controller, which can be used in a microcomputer to control the position of a servo hydraulic cylinder. In the study, a bell-shaped membership function and Takagi-Sugeno fuzzy reasoning type rules are used. The learning data are trained off-line by the back propagation method, and the normalized factors Ge and Gν are determined on-line using a modified performance index.