Research on Parameter Identification of PMSM based on EKF

The extended Kalman filter has been widely used in the parameter identification of permanent magnet synchronous motors (PMSM). In order to identify inductance, magnetic chain and other parameters at the same time, the fourth-order matrix is mostly used for multiplication and inverse operation, which takes up more computing resources and greatly affects the rapidity of the identification in practical applications. In this paper, the extended Kalman filter is downgraded to decompose the fourth-order matrix equation into two second-order equations, and a two-thread identification model is established to identify the magnetic chain and the cross-axis inductance of the permanent magnet online at the same time. The simulation results show that the improved algorithm improves the recognition speed under the premise of guaranteeing the recognition accuracy.