Adaptive robust nonlinear control of a magnetic levitation system

This paper proposes an adaptive robust nonlinear controller for position tr acking problem of a magnetic levitation system, which is governed by an SIS O second-order nonlinear differential equation. The controller is designed in a backstepping manner based on the nonlinear system model in the presenc e of parameter uncertainties. At the first step, a PI controller is designe d to stabilize the position error of the levitated object. Then at the seco nd step, an adaptive robust nonlinear controller composed of an adaptive fe edback linearization control term and a robust nonlinear damping term is de signed, to attenuate the effects of parameter uncertainties. The combined a daptive and robust approach helps to overcome some well-known practical pro blems such as high-gain feedback of the robust controller, and poor transie nt performance of the adaptive controller, so that better control performan ce can be achieved compared to the case where either is employed alone. Exp erimental results are included to show the excellent position tracking perf ormance of the designed control system. (C) 2001 Elsevier Science Ltd. All rights reserved.