Unbalance Vibration Compensation of Magnetic Bearing Systems Based on Beetle Antennae Search Algorithm

For rotating mechanical systems, it makes great sense to eliminate the vibration to improve the accuracy and stability. The synchronous centrifugal force caused by unbalance mass distribution is the main source of vibration. This paper proposed a strategy to reduce the vibration in active magnetic bearing (AMB) systems based on Beetle Antennae Search (BAS) algorithm. Firstly, a rotor levitated by two radial AMBs is modeled as dynamic differential equations. By analyzing the structure of centrifugal force, the form of compensating current is obtained. Then, in order to identify the coefficients of compensating current, the BAS algorithm is applied. During the iteration, the coefficient vector becomes the output of optimization. Next, simulations based on the dynamic model and BAS are carried out. The performances of iteration with different parameters are analyzed. The strategy can identify the parameters and suppress vibration in various initial phases of centrifugal force. The convergence rate relies on the rotating speed and the initial searching step size. After injecting the compensating signal, the vibration is reduced by over 99.7%. Random disturbance is added to the rotor and the iteration executes when disturbance operates. The strategy can also reduce the vibration when disturbance operates, which verifies the anti-disturbance ability. The results of simulations confirm that the strategy can reduce the vibration effectively and the accuracy with the convergence is realized at the same time.