LQR Optimization of an Eight-Pole Radial Active Magnetic Bearing System

Abstract

AbstractActive Magnetic Bearings (AMBs) have a powerful characteristic which, if they are correctly controlled allow a frictionless relative motion between the rotor and the stator, and grant a high-speed rotational motion without contact. However, because of the active control feature which is an energy-consuming strategy, we introduced different control current scheme in order to minimize the magnetic bearing’s energy consumption without altering the system dynamics using an appropriate controller. In this paper, we investigated the stabilization of a magnetically levitated rotor. Simplified models of current biased and non-biased radial active magnetic bearings were presented, where eight electromagnets were powered by the bias current and/or its respective control current. The magnetic loads acting on the rotor are given by linearising the relation between the coil current and the displacement around the equilibrium position to adapt the system for linear control laws. Optimal linear quadratic Riccati (LQR) controller is used. The stabilization of a high-speed rigid body rotor supported by eight magnetic bearings is studied. The model is used for control design, stability, and optimization analysis. This application tends to deal with gyroscopic coupling which should be considered in the control design. This coupling could cause system instability. Simulation results including plots of the power consumption and the dynamic response of the rotor were presented in order to judge the performances of this controller and its current consumption optimization.KeywordsActive magnetic bearingsLQR controlOptimizationBias current