Stabilization and performance of a balance beam suspended on magnetic bearings

Uses an existing experiment setup at the University of Virginia to study several fundamental issues related to the control design for active magnetic bearing systems. Our theoretical study is based on a linearized model. The model is an unstable system subject to both state and input constraints. We first determine the largest possible stability region and then construct feedback controllers that achieve stabilization in this region. We also study several performance issues such as convergence rate and disturbance rejection. The relationship between these performances and the biasing level is also revealed. Experiments are carried out on our beam balancing test rig with controllers designed for the linearized model. Experimental data not only verifies our theoretical results but also the closeness between the physical system and the linearized model.