No abstract

The undesired synchronous vibration due to rotor mass imbalance is a main disturbance source in all rotating machinery including active magnetic bearing (AMB)-rotor systems. In the AMB-rotor system, imbalance compensation, which causes the AMB actuators to spin a rotor about its geometric axis, and automation balancing, which spins a rotor about its inertial axis, are two kinds of common control aim for the rotor imbalance control. In this study, the internal relation between the imbalance compensation and the automation balancing is analyzed and a uniform control method is proposed. With the identical control algorithm, the proposed control method can realize the automation balancing or the imbalance compensation, respectively, by switching the controller’s junction position in the original control loop. The proposed control method does not depend on the dynamic plant model, because its algorithm is based on the real-time identification for the Fourier coefficient of the rotor imbalance disturbance. In this paper, the process of identification algorithm is given in detail and all the possible junction forms of the controller are illustrated. By the simulations, the identification performances of the control algorithm are compared in the conditions with three variable factors, including the signal noise ratio (SNR), the imbalance phase and the identification delay time. The noise level has considerable influence on the identification precision, but the imbalance phase has little. To prolong the identification delay time will be of benefit to improve the identification precision but slow down the identification process. Experiments, which are carried out on an AMB-rigid rotor test rig, indicate that by switching the junction position of the controller in control loop, both kinds of rotor imbalance control can achieve the good effectiveness.

No abstract