Precise control of a four degree-of-freedom permanent magnet biased active magnetic bearing system in a magnetically suspended direct-driven spindle using neural network inverse scheme

Abstract The capacity of improving the control accuracy and dynamic performance of a four degree-of-freedom (DOF) permanent magnet biased active magnetic bearing (PMBAMB) system is critical to developing and maintaining a high precision application in a magnetically suspended direct-driven spindle system. The 4-DOF PMBAMB system, however, is a multivariable, strong coupled and nonlinear system with unavoidable and unmeasured external disturbances, in addition to having parameter variations. The satisfactory control performance cannot be obtained by using traditional strategies. Therefore, it is important to present a novel control scheme to construct a robust controller with good closed-loop capability. This paper proposes a new decoupling control scheme for a 4-DOF PMBAMB in a direct-driven spindle system based on the neural network inverse (NNI) and 2- degree-of-freedom (DOF) internal model control method. By combining the inversion of the 4-DOF PMBAMB system with its original system, a new pseudolinear system can be developed. In addition, by introducing the 2-DOF internal model controller into the pseudolinear system to design extra closed-loop controllers, we can effectively eliminate the influence of the unmodeled dynamics to the decoupling control accuracy, as well as adjust the properties of tracking and disturbance rejection independently. The experimental results demonstrate the effectiveness of the proposed control scheme.