Magnetic Bearings Control Using Fuzzy Logic and a Single Neuron

Linear controllers cannot get satisfactory results for active magnetic bearings (AMB) due to the highly nonlinear and open-loop unstable electromechanical dynamics. This paper proposes a combined controller which integrates a fuzzy logic controller (FLC) with a single neuron controller for AMB. The FLC operates with a small single neuron control component when the system error is big; the FLC and the single neuron controller operate simultaneously when the system error is small. The proposed controller overcomes the defect of the FLC which has a steady state error, and the defect of a single neuron controller which has unsatisfied dynamic performance. Simulation results demonstrate that the proposed controller has improved the dynamic performance and the disturbance rejection ability of AMB.