Optimal fuzzy control of radial active magnetic bearing systems

Abstract

In this paper, a neural-fuzzy approach to develop the optimal fuzzy control of a highly nonlinear radial current-controlled active magnetic bearing (AMB) systems is proposed. A linear self-constructing neural fuzzy inference network is proposed to neural-fuzzy modeling the radial current-controlled AMB system first. Then, the corresponding optimal fuzzy control design scheme is obtained to stabilize the AMB system with minimize current consumption. Simulation results show that the proposed optimal fuzzy controller can operate in widely range of shaft position.