GA-Based Autonomous Compensator Design of 2-Degrees-Of-Freedom Positioning Controller

Abstract

This paper presents a novel Genetic Algorithm (GA) based autonomous compensator design for the fast-response and high-precision positioning of mechatronic systems. The positioning system is mainly composed of a robust 2-degrees-of-freedom (2DOF) controller based on the coprime factorization description. The feedback (FB) compensator in the 2DOF controller ensures the robust nature against the variations of load mechanical parameters. The feedforward (FF) compensator, on the other hand, allows the positioning to be fast and precise. In the controller design framework, the FB/FF compensators can be autonomously designed by GA, where both the structure and parameters of the compensators are optimized to achieve the desired robust positioning performance using μ-Analysis. The effectiveness of the proposed optimal design has been verified by experiments using a prototype of 2-mass resonant system.