Fuzzy supervisory control and GPC applied to a flexible single-link manipulator

Abstract

This paper presents the development of a classical and intelligent control approaches applied to a flexible single-link manipulator robot and compared in terms of input tracking and vibration suppression. Lagrange's equations and finite elements method are combined to compute dynamic model of a flexible link manipulator with one rigid joint. Next, dynamics are itemized to explain flexible link behavior. Then a fuzzy supervisory controller is developed and introduced in the high level of the closed-loop of the flexible manipulator. A generalized predictive controller is then developed and introduced in the flexible system closed-loop to minimize end-point residual vibrations. Simulation results obtained are compared to a Generalized Predictive Controller (GPC) in terms of end-point vibration suppression, input tracking and disturbance rejection. A conclusion encloses the paper.