Research On Unbalance Parameter Identification Of Magnetic Suspension Rotor

As a new type of high-performance support component,active magnetic bearing has the characteristics of no contact and no friction,which makes it an ideal support component for high-speed rotor.Correspondingly,when the magnetic levitation rotor rotates at high speed,the abnormal vibration of the rotor caused by its internal imbalance will be more significant.Although the active magnetic bearing can realize the active control of the rotor according to the feedback signal to a certain extent,it can not completely offset the unbalanced vibration of the rotor due to its insufficient stiffness.Therefore,the magnetic levitation rotor needs dynamic balance treatment before assembly.However,the traditional dynamic balance technology needs to disassemble and repeatedly test the magnetic levitation rotor.In view of this phenomenon,this paper proposes an unbalanced parameter identification algorithm based on the dynamic model of magnetic bearing rotor system.Firstly,taking the magnetic levitation rotor experimental platform as the prototype,the rigid dynamic model and flexible dynamic model are established.The first modes of the flexible dynamic model are verified according to the measured modal parameters of the magnetic levitation rotor.In view of the poor matching between the calculated value of the natural frequency of the rotor theoretical model and the experimental value,The flexible dynamic model of magnetic levitation rotor is optimized by genetic algorithm.Then,the established magnetic levitation rotor dynamics model is transformed into frequency domain form for parameter identification,and the unbalanced parameter identification algorithms based on rigid magnetic levitation rotor and flexible magnetic levitation rotor dynamics model are constructed respectively,and the order of flexible algorithm is reduced to solve the problem of too high order.The Simulink simulation model is built,and the unbalance and a certain proportion of Gaussian noise are added to the model to simulate the actual rotor unbalance response.According to the rotor unbalance displacement response and control current signal generated by the model,the identification accuracy of the magnetic suspension rotor unbalance parameter identification algorithm is simulated and analyzed.The influence of the accuracy of magnetic levitation rotor dynamics model on the identification accuracy of the algorithm is studied,including the influence of mass matrix,stiffness matrix,damping matrix,magnetic bearing displacement stiffness coefficient and current stiffness coefficient on the identification accuracy of the algorithm.Finally,the unbalance parameter identification algorithm of magnetic levitation rotor is verified by experiments.Six groups of unbalanced parameter loading schemes are set up,and the influence of rotor residual imbalance on the experimental results is considered.The unbalanced parameter identification algorithm of magnetic levitation rotor is experimentally studied.Experimental results show that the algorithm has good identification accuracy and proves the practicability and effectiveness of the algorithm.