Research On Displacement And Speed Sensorless Technology Of Bearingless Permanent Magnet Synchronous Motor

The traditional permanent magnet synchronous motor has the advantages of high power density,high efficiency and no excitation winding,which is an indispensable equipment in modern industrial production.The bearingless permanent magnetic synchronous motor(BPMSM)has the advantages of both magnetic bearings and traditional permanent magnet synchronous motors,which makes it have a wide application prospect in high speed and high precision equipment.The precise detection of rotor displacement and speed/position is the key to the stable suspension and rotation when the BPMSM is in operation.Generally,the displacement and speed/position signals of the rotor are obtained by installing mechanical sensors.However,the use of mechanical sensors has brought many disadvantages to the reliable operation and development of BPMSM,for example,the installation and use of sensors limit the critical speed of the motor,increase the cost of the control system,and hinder the development of bearingless motor in the practical direction.Therefore,the research on sensorless technology of the BPMSM is carried out in this dissertation for this problem.The main work and achievements are as follows:(1)A brief introduction to the two basic structures of the BPMSM is given.Based on the analysis and summary of the interaction between the electromagnetic forces inside the motor,the principle of radial suspension force generation is expressed,the BPMSM radial suspension force model and the torque mathematical model are derived.(2)For solving the problems caused by mechanical displacement sensor,the displacement sensorless technology of the five-degree-of-freedom BPMSM is studied.Combining the neural network with the theory of left inverse observation,the method for detecting displacement of two-degree-of-freedom BPMSM based on neural network left inverse is presented,and the design principle of the displacement observer based on neural network left inverse is expounded.In order to improve the accuracy of displacement observation,a flux observation method based on generalized second-order integrator is proposed.In addition,a rotor displacement self-detection model of the three-degree-of-freedom hybrid magnetic bearing is obtained by off-line training of neural network by collecting representative current-displacement sample data.The simulation results show that the displacement observer and displacement detection model proposed in this dissertation realize the displacement sensorless detection of the five-degree-of-freedom BPMSM.(3)For the study of speed sensorless of the BPMSM,a speed/position detection method based on fractional phase-locked loop is proposed.First,the back EMF observer is constructed to observe the back EMF of the torque winding.Then,the observed back EMF is taken as the input of the fractional phase-locked loop,and the rotor speed and position information is extracted from the back EMF by using the function of the fractional phase-locked loop.Finally,the feasibility of the detection method is simulated and verified.The results show that the proposed method can provide rotor speed/position information for BPMSM control with high accuracy.(4)The software and hardware of the motor digital control system are designed according to the characteristics and control requirements of the two sets of windings in the BPMSM,the torque and suspension forces of the BPMSM are controlled by two PWM waves generated by the event manager in the DSP.Based on the designed control platform,dynamic and static suspension experiments are carried out to verify the feasibility of the control system.Finally,the realization scheme of BPMSM sensorless control is given.