Linear parameter‑varying control for active magnetic bearing supported rotor systems

Abstract

Model-based control laws play an important role in stabilizing and guaranteeing high performance of active magnetic bearing (AMB) suspended high-speed machines. In the case of rotating machines with significant gyroscopic effects, the modes are separated into forward and backward whirling modes. This often leads to a situation where a single controller with fixed parameters cannot stabilize the full operational region. Addressing the challenges arising from rotating machines with significant gyroscopic effects, the application of –based linear parameter varying (LPV) control, scheduled as a function of rotational velocity, is considered in this paper. This approach is combined with an unbalance rejection strategy, where the notch parameter matrix is also updated as a function of velocity. The performance and stability of the MIMO closed-loop control are evaluated using generalized Nyquist diagrams, along with damping ratio and disk margins analysis. Such an analysis will serve as a unified framework to support the analysis as well as design and compare of LPV-based closed loop controllers. Validation of the proposed control approach is verified with experimental electrical machine equipped with AMBs.