An Adaptive Control System for an Accelerating Rotor Supported by Active Magnetic Bearings under Unbalance Disturbances

Abstract

This thesis describes an adaptive control system capable of rejecting unbalance disturbances for an accelerating rotor supported by active magnetic bearings (AMBs). The control of accelerating rotor during start-up and shut-down for rotating machines with AMBs is di - cult, especially when the rotor has strong unbalance, which results in sinusoidal disturbances of time-varying frequency and time-varying amplitude. A stabilizing controller alone is not able to obtain a satisfying dynamic performance in this situation. Therefore, unbalance disturbance rejection control is needed. The research work on this topic is mostly based on steady-state rotor model, therefore, those methods are not applicable during rotor acceleration. In this thesis, a transient rigid rotor model is rst developed and then used for controller design and stability analysis. The adaptive control system, which falls into the category of model reference adaptive system (MRAS), includes a PD-typed controller for stabilizing the system, and an adaptive disturbance rejection (ADR) controller for rejecting disturbances. The ADR controller utilizes rotor speed measurement, which is a piece of known information for most rotating machines, to construct a feedforward regressor vector, and makes the rejection of unbalance disturbances during rotor acceleration possible. It is proved by Lyapunov theory that the adaptive control system is asymptotically stable. Dynamic performances of the adaptive control system are shown by computer simulation.