Fuzzy and Online Trained Adaptive Neural Network Controller for an AMB System

Abstract

In this paper, an intelligent control method to positioning an active magnetic bearing (AMB) system is proposed, using the emerging approaches of the fuzzy logic controller (FLC) and on-line trained adaptive neural network controller (NNC). An AMB system supports a rotating shaft, without any physical contact by using electromagnetic forces. In the presented controller system, the FLC is first designed to identify the parameter of an AMB system. Then the initial training data with two inputs, the error and derivate of the error, and one output signal from FLC can be obtained. Finally, a NNC with on-line training features is designed by using S-function in Matlab/Simulink to achieve better performance. The results of the AMB system indicated that the system responds with satisfactory control performance, without a small overshoot, and obtains better transient and steady-state responses under various operating conditions. Moreover, this method has been experimentally verified on the AMB system. The proposed controller can be feasibly applied to AMB systems exposed to various external disturbances, demonstrating the effectiveness of the NNC with self-learning and self-improving capacities is proven.