Adaptive Forced Balancing for Multivariable Systems

Abstract

Active magnetic bearing (AMB) actuators support rotors without friction but require feedback control for stabilization and performance. Autobalancing compensation causes AMBs to spin a rotor about its inertial axis to eliminate synchronous force transmission from mass unbalance. Because mass unbalance constitutes a sinusoidal sensor disturbance within the bandwidth, conventional methods can either cause instability or fail to preserve desired bandwidth. We introduce a new method called adaptive forced balancing (AFB) which overcomes these problems. We consider AFB with a frequency tracking capability for SISO systems (i.e., single-end AMB suspensions) and show how to extend it for the MIMO case as applied to a double-end AMB suspension.