Identification Of The Magnetic Bearings Support Parameters Using An Extend Kalman Filter

As a new type of support bearings, magnetic bearing have several advantages incomparable totraditional bearings, and has been applied to many fields. Due to the stiffness and damping parametersare adjustable online, the supporting characteristics of magnetic bearing has a more importantinfluence on the dynamic characteristics of a system. Therefore, identifying the supporting parametersof magnetic bearings is an important component of maglev technology research, but most identifyingmethods were based on the calculation in frequency domain. This article mainly studied on stiffnessand damping parameter identification of magnetic bearing by Extend Kalman Filter (EKF) based ontime-domain calculation as it was excited by imbalance or impact excitation.Firstly, illustrated the principle of the EKF identifying method based on nonlinear system,considered the structure of the magnetic bearings–rotor system in this paper, set the magnetic bearingstiffness and damping parameters being system state parameters, and ensured the process of EKFidentifying method. Then, obtained the displacement response of the rotor under unbalance excitationwhen steadily working and used the Matlab program to calculate the supporting parameters,separately in simulation and experience. And at the same time, identified the supporting parameters byboundary element method and compared with the EKF methods.Built up a magnetic-rotor experiment rig for stiffness and damping identification under impactexcitation and simulated in Samcef. Changed the excitation force to transient and changed input termof EKF algorithm. By signal collection and data processing, the displacement response of the rotorunder the hammer impact force can be obtained. At last identified the supporting parametersseparately by the EKF method and transfer matrix method and compared the two methods.