Robust predictive current control of permanent magnet synchronous motor using voltage coefficient matrix update

Abstract

In this study, a novel robust deadbeat predictive current control algorithm using voltage coefficient matrix update (DPCC-VCMU) is proposed for permanent magnet synchronous motor (PMSM), which can eliminate the influence of model parameter mismatch. First, the parameter sensitivity of conventional deadbeat predictive current control (DPCC) is discussed, and the multiple parameter error online identifiers are designed based on an improved Adaline neural network (ANN) weight adjustment algorithm. The novel identification structure with parameter mismatch error as the weight of neural network is then proposed to enhance the current tracking robustness under model parameter mismatch. Moreover, the parameter error identification value is used to update the coefficient matrix of the control voltage and predict the current value of next moment simultaneously. Finally, compared with conventional DPCC method, the simulation and experimental results verify that the proposed DPCC-VCMU-algorithm can significantly improve the current response accuracy by using voltage coefficient matrix update, while achieving lower torque ripple and smaller current harmonics.

Keywords

• Parameter error identification
• Model parameter mismatch
• Adaline neural network
• Predictive current control
• Permanent magnet synchronous motor