Model predictive control of induction motor based on amplitude-phase motion equation

Tedious adjustment of the weighting factor restricts the wide application of conventional model predictive torque control (MPTC) for induction motors (IMs) in practice. To solve this problem, this study introduces a new variable R (reactive torque) to control the flux of IM, which is the dual of torque T, with radial orientation and in relation to reactive power. Then, a novel MPTC strategy based on variable R and torque T is proposed and named RT-MPC. Compared with conventional MPTC, RT-MPC uses a pair of variables with the same dimensions and time scales instead of the torque and flux amplitude in MPTC. This approach eliminates the weighting factor of the cost function, unifies the time scales of control variables, and avoids prediction of rotor flux at the (k+2) instant. Simulation and experimental results show that RT-MPC can avoid the weighting factor, reduce switching frequency, and reduce control complexity. RT-MPC can also maintain excellent dynamic and steady-state performance similar to conventional MPTC. Hence, the proposed method is expected to be more practical.