Online Full-parameter Estimation of SynRM Based on the RLS and LMS Algorithm

Aiming at the rank-deficient issue of the full-parameter identification model of synchronous reluctance motor (SynRM), an online full-parameter identification scheme combined with the reclusive least square (RLS) algorithm and the least mean square (LMS) algorithm is proposed in this paper. Due to the characteristics that the inductance parameters vary rapidly with the currents, while the stator resistance varies slowly with the winding temperature, the RLS algorithm with a faster convergence rate is proposed to identify the inductance parameters quickly, and based on the RLS identification results, the LMS algorithm with less calculation is used to update the stator resistance. This method avoids the effect of using fixed parameters to identify the other parameters, and it can follow the changes in motor parameters well. Finally, the simulation results in the MATLAB/Simulink verify the effectiveness of the proposed scheme.