Modeling and observer-based nonlinear control of a magnetic levitation system

This paper focuses on the exact feedback linearization of a nonlinear plant using estimated states. It is well-known that the separation principle is not guaranteed for nonlinear plants, and appropriate hypothesis are commonly assumed to assure local stability properties. Here it is shown that the separation principle is valid for the design of observer-based state feedback of an experimental prototype of a magnetic levitation system. Once the model and its parameters are obtained, state feedback and observer design are separately and successfully carried out. Estimated states are then fed back and results show that stability is assured for perturbations in the equilibrium and estimated states.