A control method of bearingless induction motor based on neural network

To Solve The nonlinearity and coupling problem of bearingless induction motor, a decoupling control method based on neural network inverse system is proposed. Under the condition of considering stator current dynamics of torque system, the stator voltage are selected as input variables, and by neural network training, the inverse system model of torque system is identified, which is used to decouple the torque system into two second order linear subsystems, include rotor flux-linkage and speed subsystems. Then in magnetic suspension control system, the negative feedback control of radial displacement and the compensation of unbalance unilateral electromagnetic pull are achieved; the required air gap flux-linkage of torque system is calculated online according to rotor flux-linkage and stator current components. From the simulation results: compared with the decoupling control method based on analytic inverse system, the control system owns stronger dynamic response characteristics and stronger ability of anti disturbance; the proposed control strategy is effective.