The zero-bias current controlled way is proposed to cut down the power consumption of the active magnetic bearing in a power magnetically levitated spindle system. The zero-bias current controlled way is easier to realize than the zero-bias flux controlled way, since current can be detected directly, while flux is hard to be measured in practice. Besides, the active magnetic bearing suffers from lumped uncertainty including parameter uncertainty and external load, and the displacement of rotor caused by lumped uncertainty is undesirable. In practice, the upper bound of the lumped uncertainty especially the external load is hard to obtain, making it hard to choose parameters for a traditional sliding mode control. The adaptive backstepping sliding mode control method combining both the advantages of sliding mode procedure and backstepping procedure is proposed to solve this problem. Furthermore, the upper bound of lumped uncertainty is estimated in real time by an adaptive law. In this paper, first a new zero-bias current active magnetic bearing system model with lumped uncertainty is built; then two controllers based on the sliding mode control and adaptive backstepping sliding mode control methods are designed, respectively, and the stability analyses are given for the two controllers via Lyapunov function; finally, the effectiveness of the proposed adaptive backstepping sliding mode control approach for a zero-bias current active magnetic bearing system is verified by the simulation and experiment results.
A System Identification Technique Using Bias Current Perturbation for the Determination of the Magnetic Axes of an Active Magnetic Bearing
