Research On Coordinated Suppport Control Methodology In Self-sensing Active Magnetic Bearings

Compared with the traditional mechanical support,The magnetic bearings are attracted on the high ratational speeds rotor because magnetic lecitation system can provide contact free suspension without friction and lubrication.And the active magnetic bearings(AMBs)have become a key part in the industrial applications with the excellent performance of active vibration control ability.Active magnetic bearings require the closed-loop controller with positon sensor to achieve the targeted rotor performance which is effected by the instability casused by the sensors and actuators non-collocated,the unbalanced vibration of the rotor,the vibration mode and supercritical operation,uncertain disturbances,etc.Although many researchers and engineers made great efforts,it is difficult to satisfy the requirements in application.In order to improve the magnetic levitation system operation,a solution has developed by the coordinated support control system based on auto-disturbance rejection controller(ADRC)which embed by the self-sensing magnetic bearings and the self-sensing magnetic dampers collocated.In this work a self-sensing method was introduced for the estimation of position of active magnetic bearings,based on the least square identification algorithm.To extract rotor position information from coil current signal accurately,a detailed mathematical model of position estimator was derived and analyzed.Former estimators of self-sensing AMBs are mainly based on a ideal rotor position demodulator model of the first switching harmonics,so that it is difficult to improve the precision of the real demodulator.By utilizing the cosine Fourier series of the switching current and least square identification,an estimation algorithm for the inductance of AMBs was established.Meanwhile the inductance analytical vector of charging and discharging phase was described in detail,and an estimation of inductance was given by the high precise sampling current.The cascaded position controller of AMBs was served in real-time hardware on the rig.Afterwards,to regulate and control the equivalent stiffness and equivalent damping in the magnetic levitation rotor.a coordinated support model with stiffness and damping is proposed.The control systemt consists of the self-sensing magnetic bearings and electromagnetic damper,and the rotor state is evaluated in real time through the linear extended state observer(LESO).The internal influence and the external disturbances of the magnetic levitation rotor are introduced into the ADRC model,and the double closed-loop controller is constructed with LESO.Linear auto disturbance rejection controller(LADRC)can not only solve the multi-parameters tuning problem,but also evaluate and compensate the uncertain dynamic,parameter perturbation and external interference of the magnetic levitation rotor system in real time.Meanwhile the coordinated support model with stiffness and damping is improved by LADRC.Then the gyro effect and transcritical vibration characteristics of the maglev rotor are analyzed by virtual experiments.For the rotor gyroscopic effect suppression,a coordinated support control model is proposed by the cross-feedback decoupling controller which is introduced the damping,and overcomes the limitation of the time delay.Aiming at the supercritical vibration suppression,in this work the finite element and the modal analysis are adopted to simulate the supercritical vibration characteristics of the magnetic levitated rotorThe rig of five degree of freedom(5-DOF)magnetic bearings is developed based on the single degree of freedom(1-DOF)magnetic bearing device,which is based on the distributed computer control system.And the tests of rotor vibration suppression mainly include the test of position estimation of the self-sensing magnetic bearings,the test of the electromagnetic damper regulated and controlled,and the disturbance rejection test under the coordinated control of the damping and LADRC,and critical vibration suppression on the 1-DOF and 5-DOF rig.The experimental results show that the coordinated support controller based on LADRC and damping has faster convergence speed and higher control performance.To meet the requirements of high-speed,flexibility and stablity in the magnetic levitation application,this dissertation has fully tried to reduce the structure of the magnetic levitation system by the self-sensing technical,and puts forward the basic control model based on the coordinated support by stiffness and damping.This work provides a new idea and implementation method for the high-speed operation of magnetic bearings.