Design And Implementation Of A Feedback Linearizing Controller And Kalman Filter For A Magnetic Levitation System

The principal investigation undertaken in this research is the development and subsequent implementation of a feedback linearizing nonlinear controller and extended Kalman filter for a laboratory-based Magnetic Levitation (Maglev) device. The Maglev hardware is both highly nonlinear and open-loop unstable in its dynamic response. The control and estimation scheme proposed in this work is first validated using an increasingly sophisticated level of simulations. The control and estimation algorithm had the sensor noise identified using hardware data. Then, the plant noise covariance is tuned using both on-line and off-line hardware data. Additionally, the output of the nonlinear controller is then mapped to hardware-suitable levels using a digital lead-lag controller. Experimental results are included where the system is given a set of reference trajectories to track.