Identification of Magnetic Bearing Systems via Wavelet Signal Processing

• Main Authors: Chien-Ting Chen, 陳建廷
• Other Authors: Nan-Chyuan Tsai
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/16128267817929573078

碩士 === 國立成功大學 === 機械工程學系碩博士班 === 95 === Most traditional methods in system identification are based on the analysis of measured signal in either time or frequency domain. In recent years, some technologies in time-frequency domain which utilize wavelet analysis in the context of system identification were developed. The purpose of this thesis is to develop an algorithm based upon Daubechies wavelet expansions to identify unknown system parameters of the 4-pole active magnetic bearing systems (AMBs). The procedure stems from the equation of motion obtained by Daubechies wavelet and means of Least Squared Method. The wavelet identification method is compared with another populary-used method: Eigensysem Realization Algorithm (ERA). The proposed algorithms were examined by numerical simulations and experiments. The test rig is equipped with dSPACE DS-1104 and MATLAB/Simulink. The results of numerical simulations and experiments verify that the Wavelets System Identification Method has better efficacy in convergence and accuracy.