A Comparison of Linear Periodic and Nonlinear Control Strategies for Self-Sensing Magnetic Bearings

In this paper, the controller design for self–sensing magnetic bearings is considered using a linear time periodic (LTP) framework. The time periodicity comes from a deliberate bias flux excitation signal allowing to improve the plant properties. Design trade–offs using periodic controllers are investigated. Three design objectives are considered: 1) low sensitivity norm to ensure good robustness, 2) low controller gain to attenuate current measurement noise and to avoid voltage saturation, and 3) low mechanical compliance against external perturbation forces to ensure high bearing stiffness. Different LTP H8 controllers are designed using lifting techniques and linear matrix inequalities (LMI). The resulting controllers are validated using closed–loop simulations with the nonlinear plant.