Adaptive Fuzzy Kalman Filter Applied for Identification of Rotor/Active Magnetic Bearing Dynamics

Yu-Hsun Liu, Nan-Chyuan Tsai, and Hsin-Lin Chiu


Active Magnetic Bearing, System Identification, Kalman Filter, Fuzzy Logic Algorithm


The main goal of this research is to identify the system parameters of the dynamics of Rotor/Active Magnetic Bearing (Rotor/AMB) system employed in turbo molecular pumps (TMPs). The identification approach adopted in this work is based on experimental analysis and the application of Fuzzy Logic Adaptive Control-Extended Kalman Filter (FLAC-EKF). The estimation error for either system state or system parameters can be gradually converged to zero via self-tuning of the design parameters of FLAC-EKF. The proposed algorithm has been verified by numerical simulations and intensive experiments. It is concluded that the FLAC-EKF can exhibit satisfactory performance in terms of estimation accuracy on the system parameters even under contamination of a certain degree of process disturbance and sensor noise.

Important Links:

Go Back