Sliding mode control with online fuzzy tuning: application to a robot manipulator

Abstract

This paper presents an approach to improve the sliding mode controller performance. Accordingly, a fuzzy controller is designed based on a limited number of rules and independent of the sliding mode controller complexity. To enhance the performance, this controller continuously optimize the sliding mode controller parameters including hitting control gain, boundary layer thickness, sliding surface slope and intercept. The controller is applied to a two-degree-of-freedom robot and the experimental results are compared with a QFT based and a PID controller. The results demonstrate the significant performance improvement of the proposed controller.