Servocontroller tuning with fuzzy logic

Abstract

A fuzzy-logic-based self-tuning proportional-integral (PI) controller is designed that achieves robust tracking in unknown plants which may be open-loop unstable. Tuning of the conventional servocontroller parameters is accomplished through fuzzy approximate reasoning techniques. A Lyapunov function is used as a performance index in formulating the fuzzy rules. The proposed controller parameter tuning scheme greatly reduces the dimension of the fuzzy rule set. Computer simulations on an unstable nonlinear plant illustrate the validity of the approach. >