An Adaptive Multi-Objective Controller for Flexible Rotor and Magnetic Bearing Systems

Abstract

This paper considers two issues in the vibration of rotating machines, namely, control of rotor vibration and control of the forces transmitted to the base. An adaptive multi-objective method is developed to tackle these issues simultaneously using magnetic bearings. A two-stage weighting strategy is developed, involving base weightings calculated by using a singular value decomposition of the system’s receptance matrices and an adjustable single weighting parameter tuned automatically to shift the balance between the two objective functions in accordance with performance specifications. This new real-time controller does not require measurements in addition to those required for the open-loop adaptive strategy. Frequency and time domain simulations of an existing experimental rig are used to assess the effectiveness of the proposed multi-objective adaptive controller as a precursor to an experimental study.