| Compared to traditional contact bearings,the magnetic bearing is a new type of noncontact bearing that uses electromagnetic force to achieve stable rotor levitation,with low noise,no wear,no lubrication and controlled stiffness and damping.The radial magnetic bearing in its basic configuration suffers from localized magnetic force line clustering,which tends to exacerbate magnetic leakage,losses and temperature rise;at the same time,it can be improved for traditional magnetic levitation bearing PID controller.In this paper,a support system for a flywheel battery based on a magnetic bearing is designed,the structural design and electromagnetic characteristics of the radial magnetic bearing and the axial magnetic bearing to investigate the support characteristics are down on the basis of a traditional magnetic bearing PID controller,and combine it with a radial basis function neural network to optimize the control effect.(1)According to the structural characteristics of the on-board flywheel battery,a support system scheme was designed,a modal analysis of the flywheel rotor system was completed,and the form and selection of the magnetic bearings and auxiliary bearings were determined.A simplified mechanical model of the magnetic bearing was established,the mathematical equation of electromagnetic force was derived,and the principle of differential control of the magnetic levitation bearing was analyzed.(2)The structural design of radial magnetic bearings and axial magnetic bearings was completed,including the selection of the number of magnetic poles,the calculation of the specific structural parameters of the stator and rotor of the magnetic bearings,etc.The electromagnetic characteristics of radial magnetic bearings and axial magnetic bearings were analyzed with the help of ANSYS Maxwell.The electromagnetic characteristics of radial magnetic bearings and axial magnetic bearings were analyzed with the aid of ANSYS Maxwell electromagnetic analysis software.The effects of different excitation methods on the coupling between the magnetic poles of the radial magnetic bearing and the effects of different modes of energization on the distribution of magnetic lines in the radial magnetic bearing when the coil winding was loaded with bias current are also analyzed.In view of the shortcomings of the basic structure of the radial magnetic bearing,structural improvement and optimization solutions were designed to compare the electromagnetic characteristics of the radial magnetic bearing under different pole and slot schemes.The improved pole scheme effectively alleviated the edge effect problem of the stator pole.The elliptical transition angle and circular transition angle solutions were used to achieve a smooth transition at the stator slot inflection point and to reduce the maximum magnetic induction at the pole by 8.37% and 11.0%respectively,effectively improving the distribution of magnetic lines of force.(3)A single-degree-of-freedom PID control model of the magnetic bearing was established,and the influence of the control parameters on the analysis of the magnetic bearing support characteristics was analyzed according to the constrained control parameter range based on the Rouse stability determination,and the experimental simulation of the magnetic bearing heave was completed with the help of MATLAB/Simulink.On the basis of the conventional PID controller for magnetic bearings,a radial basis function neural network was combined with the controller design to realize the adaptive dynamic adjustment of the control parameters of the magnetic bearing controller,and the analysis of the rotor response under the step signal and sinusoidal signal excitation of the improved magnetic bearing controller was completed. |