A Parameter Identification Method Based on Reduced-Order Observer for SPMSM With Deadbeat Predictive Current Control

Deadbeat predictive current control (DPCC) is widely adopted in surface-mounted permanent magnet synchronous motors due to its superior control performance. However, DPCC heavily relies on the accuracy of the motor mathematical model, which means that parameter mismatches can induce current errors and pulsations. To enhance the parameter robustness of DPCC, this article proposes an online parameter identification method based on a reduced-order observer. First, sensitivity analysis of stator resistance, stator inductance, and rotor flux linkage is conducted to investigate current errors caused by individual parameter mismatches. Subsequently, stator inductance is estimated based on voltage equations, with recursive least squares. A reduced-order state observer is then employed to estimate rotor flux linkage, accompanied by stability analysis of the observer. Finally, the current control performance and parameter identification effectiveness of DPCC with the proposed method are analyzed by simulations and comparative experiments. It is verified that the proposed method effectively suppresses steady-state current errors and harmonic components induced by parameter mismatches.