Multi-Model Robust LQ Control of an Active Magnetic Bearing

The problem of designing a robust LQ controller for a multi-model active magnetic bearing (AMB), is tackled. The state variables are regulated. A tangent linearized model of the AMB near an equilibrium point, is considered, and a minmax optimal controller is designed for an uncertain model of the AMB with parameters in a known set. This set of parameters defines three models of the AMB, and the designed controller guarantees sufficiently "good" performance for each of the three models. The robust LQ controller is compared through simulations with the traditional finite time LQ controller and a mixed sensitivity H_infin robust controller, for an uncertain system.