A study on the extended horizon adaptive control algorithm of the magnetic support system for high-speed rotor machines

Abstract

The paper focuses on research studies regarding the properties of active magnetic bearing supporting system when it is implemented in a high-speed rotating machine. The laboratory active magnetic suspension system under consideration consists of a massive, single-disk rotor-shaft supported by two radial and one axial (thrust) magnetic bearing. Here, the most important analysis concerns control issues in the active magnetic bearing when it uses the Extended Horizon Adaptive Control. This control method was chosen for deep analysis to reduce classical controllers’ weaknesses, such as the steady-state error of the proportional-derivative method and the saturation effect of integral operation in the proportional-integral-derivative approach. The control system’s detailed theoretical and experimental analyses are compared to the proportional-derivative algorithm, using time-histories of the vertical rotor-shaft displacement and the control current time-histories at zero, variable and constant rotational speeds. The obtained sufficiently good quantitative and qualitative agreement of the registered theoretical and experimental results confirmed the reliability of the analytical fundamentals of both methods used to control active magnetic bearings, as well as the significant advantages of the predictive Extended Horizon Adaptive Control method compared to the classic proportional-derivative approach.