An Improved Model Predictive Current Control of Five-Phase PMSM Drive for Torque and Flux Ripple Alleviation

Abstract

The model predictive current control (MPCC) is the popular technique for five-phase permanent magnet synchronous motor (FP-PMSM) drives. However, the implementation of the MPCC algorithm is difficult for a five-phase two-level voltage source inverter fed PMSM drive due to more switching states and an increased number of calculations. The conventional MPCC (CMPCC) methods involve larger computations and limited control over third-order harmonics. To address these issues, an improved MPCC method is presented for the FP-PMSM drive using preselection of voltage vectors. In this scheme, from the previous sample voltage vector information, a five-voltage vector group is formed. One of the voltage vectors in the group is specially intended to reduce the harmonic currents. The five grouped vectors are used for predicting the stator currents and cost function optimization. The optimal voltage vector among these five vectors is determined and applied in the next sample period to achieve the control of currents and subsequently the torque and flux in fundamental and harmonic subspaces. The proposed scheme is experimentally verified to determine its efficacy in reducing flux ripple, torque ripple, and computational burden and compared the obtained results with the CMPCC schemes and other MPCC schemes reported in the literature recently.