Research On Inverse-control Of Single-winding Bearingless Switch Reluctance Motor Using Relevance Vector Machine

With the development of the social industry,the demand for the performance of the motor is increasing.It is very difficult to reach the requirement of super high speed for the motor of the machine with bearing.Bearingless switched reluctance motor by adding maglev technology in the switched reluctance motor.Efficiency conversion of mechanical and electrical,small axial space and instant speed are the motor's advantages.which is applied to flywheel energy storage,machinery manufacturing,intelligent equipment and other fields.Single-winding bearingless switched reluctance motor(SWBSRM)changes bearingless switched reluctance motor's structure with simple structure,reliable operation,easy control,etc.Which has been extensively researched.Although the single-winding bearingless switch reluctance motor has many advantages,but it is a system with non-linear and strong coupling characteristics.There is a coupling serious between the torque and the suspended force.The traditional inverse decoupling method depends on mathematical model accurately of the single-winding bearingless switched reluctance motor.Aiming at this problem,a new inverse model decoupling control method is proposed,which is based on a 12/8 pole SWBSRM.The feasibility and superiority of the method are verified by simulation experiments.Which is a foundation of the motor in the practical application.There are many research contents and achievements :(1)The 12/8 pole SWBSRM's operation principle is analyzed,and its advantages are compared with the double-winding bearingless reluctance motor.Based on the principle of virtual work and the equivalent magnetic circuit method,the mathematical model of the radial force and torque of the single-winding magnetic reluctance switched reluctance motor are deduced.The accuracy of the mathematical model is verified by finite element analysis.The results show that the constructed mathematical model accurately reflects the motor's characteristics.(2)The finite element simulation is used to analyze the influence of the parameters of the 12/8 pole SWBSRM on the performance.The key parameters are used to optimize by the particle swarm optimization,and the maximum of the average suspended forceand the average torque are the optimized target,which's specific optimization process be described.last,The finite element analysis results show that the motor performance has been improved greatly after parameter optimization.(3)In view of the coupling problem of the 12/8 pole single-winding bearingless reluctance motor,the reversibility of the model is verified,and the inverse system model is constructed by the relevance vector machine.Before the original system,the inverse system connect with the original system,and the nonlinear system changes into a pseudo-linear system.Feed-back control system achieve accurate decoupling control of the motor's torque and the radial displacement of rotor.Simulation show that the decoupling effect is excellent and the anti-interference ability of the system is strong.(4)A digital experiment system based on DSP and FPGA is designed.The structure of the control system is studied firstly.Than,The power converter,signal detction and the communication interface between DSP and FPGA are designed.Those work is a foundation for the next experiment.