Model Modulated Speed and Current Predictive Control (M2PC) of Six-Phase Induction Motors including Magnetic Saturation and Iron Losses

The dynamic modeling of dual three-phase induction motors is very complex and, at the same time, crucial because it is the starting point for devising the related control strategies, traditionally Field Oriented Control (FOC) and Direct Torque Control (DTC). The control system theory, however, offers an important corpus of nonlinear control methodologies, among which the predictive control (PC), including the so-called model predictive control (MPC). MPC can be further divided into two sub-categories: MPC with continuous control set and MPC with finite control set. This paper proposes a model-modulated predictive controller (M2PC) applied to a dual three-phase induction motor. It tries to improve all the previous contributions in the following original aspects. Here MPC is exploited for both current and speed control. In particular, the speed prediction is performed by adopting the mechanical equation of the motor and an original load torque estimator based on an unknown input observer (NUIO). Moreover, as for both the current and speed predictions, a so-called modified Euler discretization process has been adopted, permitting better accuracy in the state predictions compared to the classic Euler method. Furthermore, the proposed MPC is based on an original dynamic model of the three-phase induction motor, which can be straightforwardly used also in case of machine with an unbalanced structure, e.g. faulty motor with any stator phase winding. The proposed MPC can be thus easily upgraded to make the drive work in case of opening of any phase of the motor. The proposed MPC, synthesizes the inverter voltage vectors so as to fulfill the stator current requirements in the sD − sQ subspace and minimize the current components in the z1 − z2 subspace, but it exploits also the medium voltage vectors, besides the large ones. The proposed MPC has been implemented in numerical simulation and experimentally tested on a suitably developed test set-up.