Experimental evaluation of model predictive current control of a five-phase induction motor using all switching states

Abstract

Finite-control-set model predictive control (FCS-MPC) of a five-phase induction motor is investigated in the paper. The predictive controller is operated in the rotor flux oriented reference frame and the predictive current control scheme includes both d-q flux/torque producing current components and non-flux/torque producing x-y current components in the cost function to minimise additional winding losses. All the available switching states (32 in total) are included in the optimization process. Impact of the x-y currents' weighting factor (W xy ) and stator fundamental frequency on the average values in both the primary and the secondary plane and on the phase current ripple are investigated using simulation. Next, the predictive control scheme is further verified experimentally. It is shown that the FCS-MPC regulates the axis currents instantaneously following reference changes, without any disturbance in other axis currents.