On Instrumental Variable-Based Method for Identification of Permanent Magnet Synchronous Machine by Noisy Data

The identification of parameters of the permanent magnet synchronous machine (PMSM) is considered. The paper uses a discrete model of the permanent magnet synchronous machine. The proposed algorithm makes it possible to consider the errors in measurements of current, voltage, and electrical angular speed. Extended instrument variables are used to identify parameters. To solve the overdetermined system, total least squares are used. This makes it possible to improve the accuracy of estimates of extended instrumental variables on a finite sample. The simulation results confirmed the high accuracy of the resulting estimates.