Research On Nonlinear Control System Of A Bearingless Permanent Magnet Synchronous Motor

Bearingless permanent magnet synchronous motor(BPMSM) is a new motor withrotating and self-levitating rotor. With no friction, no abrasion, no lubrication andsealing, high precision, low maintenance cost, and without excitation current, it iswidely used in chemical pumps, worm molecular pumps, blood pumps, high-speedmilling machines, compressors and a high speed flywheel and other equipment.Compared with ordinary PMSM, the rotor of BPMSM is self-levitating when it worksat a steady state. However, the rotor position, speed and radial displacement are moreprone to vibration and mutation because of the external disturbance, parameterperturbation and other factors. This paper focuses on the mechanism of BPMSMsuspension forces and has in-depth theoretical analysis and experimental study, suchas the mathematical model, nonlinear decoupling control, and position sensorlessoperation.Illustrated by the case of BPMSM with surface-mounted four pole torquewinding and two pole suspension winding, Based on the Maxwell tensor, in-depthanalysis of the levitation force generating mechanism was done. Considering thepermanent magnet torque winding, suspension winding, rotor eccentricity and otherfactors, the analysis and the establishment of the suspension force mathematical modeland mathematical model of electromagnetic torque were built. Meanwhile, startingfrom the rotor dynamics, the equations of motion of the rotor and system were set up.Based on nonlinear decoupling of rotor field oriented control, suspensioncontrol system uses Conventional PID for controlling winding radial suspension forceand uses PI for controlling speed. But the dynamic performance becomes bad whenthe load disturbance happens. Therefore, for the first time the fractional PIλBPMSMcontrol will be applied to the controller. Fractional controller has a lot of superioritythat the integer order system is unable to achieve. Simulation results show that thecontrol system using BPMSM fractional PIλhas a faster response speed, betterrobustness than using integer order controller PI control.In order to solve the buffeting problem of the system when parameters changedor disturbance generated, this paper design a fractional sliding mode controller basedon neural network to control suspension radial suspension force windings to reducebuffeting vibration. Simulation and experimental results show that the neural network fractional sliding mode control system can not only reduce buffeting, but also achievehigh integrated control performance. The results are useful for the application of thefractional order controller to the BPMSM suspension control system，and establish asolid foundation for developing the high-speed operation and stable suspension ofBPMSM.According to the requirements for the BPMSM sensorless position operation,This paper proposed BPMSM sensorless position operation control base on slidingmode observer and verified the above control system by experiment test.Finally, the research and the innovation of this dissertation are summarized, andfurther suggestions for the research are proposed.