Deadbeat Predictive Current Control for SPMSM at Low Switching Frequency With Moving Horizon Estimator

Abstract

The deadbeat predictive current control (DPCC) of permanent-magnet synchronous motors (PMSMs) is well known for its simple structure and fast dynamic. However, it is sensitive to parameter mismatches and system delays, which becomes more serious at the low switching frequency. To handle such problems, this article proposes an accurate deadbeat control method based on an exact model of surface-mounted PMSM in a stationary two-phase frame with a moving horizon estimator (MHE). The motor model is obtained from differential equations, and transformations between stationary frame and rotating frame are reduced to eliminate the influence of the lag of electrical angle. MHE is utilized to estimate the back-electromotive force (EMF) and disturbance caused by parameter variations. The rotor flux is eliminated in the algorithm and the resistance and inductance mismatches are compensated by the estimator. Several simulations and experiments are performed in three methods-the conventional DPCC in rotating $dq $ -axis, DPCC in $\alpha \beta $ -axis with the accurate model, and the proposed method. These comparisons validate that the proposed method has good current tracking performance at low switching frequency and a strong robustness to parameter variations.