Neural robust control of a high-speed flexible rotor supported on active magnetic bearings

Abstract

This paper presents an methodology for designing robust controllers for active magnetic bearings (AMBs) which benefits from uncertainty identification using artificial neural networks (ANNs). A high-speed flexible rotor supported by AMBs is modeled using analytical approaches, finite element analysis, and system identification. ANNs learn the statistical bounds of model uncertainty resulting from unmodeled dynamics and parameter variations. These bounds are incorporated into the synthesis of multivariable robust controllers. Experimental results on an AMB test rig reveal the benefits of this combination of intelligent system identification and robust control: significant performance improvements vs. conventional robust control in the face of process disturbances.