Identification of MIMO magnetic bearing system using continuous subspace method with frequency sampling filters approach

This paper presents a system identification of multi input multi output (MIMO) magnetic bearing system. It comprises of two steps. The first step employs a MIMO frequency sampling filters, in which, the raw data will be analyzed in order to obtain a meaningful multivariable parameters that describe the empirical model of the analyzed data. Then, the information of the empirical model will be used in the second step, in which, the MIMO continuous time subspace method will be used to obtain the state space model of the system. The application based on magnetic bearing system apparatus data will be used to demonstrate the efficacy of the developed MIMO algorithms and the approach used. Based on result, the MIMO two steps identification provide significant performance as compared to direct identification approach.