Multiparameter Identification for SPMSMs using NLMS Adaptive Filters and Extended Sliding-Mode Observer

The authors propose a parameter identification method for sequential identification of electrical and mechanical parameters of surface-mounted permanent magnet synchronous motors (SPMSMs). Two normalised least mean square (NLMS) adaptive filters (AFs) are designed for identifying the electrical parameters, where the first AF identifies the stator inductance and the second AF identifies the stator resistance and rotor flux linkage. The NLMS AFs achieve faster transient responses than recursive least squares (RLS) AFs owing to lower computing load and smaller memory size. Regarding mechanical parameters, an extended sliding-mode mechanical parameter observer (ESMMPO) is employed to estimate the system disturbance and angular velocity, from which the rotational inertia, viscous damping coefficient, and load torque are identified. The rotor flux linkage identified from the second NLMS AF is used for estimating the real-time system disturbance of the ESMMPO, which enables the identification of mechanical parameters with higher accuracy. The proposed method effectively integrates the NLMS AFs and ESMMPO into a single framework to identify both electrical and mechanical parameters of the SPMSMs. The experimental results of the proposed method are compared with those of RLS AFs and the conventional ESMMPO, which demonstrates the faster response and less steady-state parameter errors of the proposed method.