Sampled data robust control for a magnetic levitation system — Some practical considerations

Abstract

This paper presents a sampled data robust control design for a magnetic levitation system with unstructured uncertainties; which is the case in many industrial applications due to wear and tear and un-modelled dynamics. Since the industry is still inclined towards classical control design, the performance of a sampled-data H-infinity loop shaping controller has been presented where the weights have been directly taken from a lead-lag compensator. We mainly demonstrate on a real system that the presented approach provides a robust solution with good performance by taking the advantage of the state of the art robust control algorithms; while the workload of the design is almost the same as classical approaches. Both classical and robust controllers are tested on a lab-based magnetic levitation system having nonlinear dynamics with unstructured uncertainties.