Application of cascaded constrained receding horizon predictive control to an induction machine

Cascaded predictive structures have been successfully implemented for the velocity and position control of motor drives, with good temporal and frequency results (small overshoot, cancellation of disturbances, cancellation of position and tracking errors, good stability margins...). A new cascaded predictive structure is thus presented, which develops in the inner loop a constrained receding horizon predictive control algorithm with multiple reference models (CRHPC/MRM), and in the external loop a classical generalized predictive control version (GPC). The interest of such a structure is then examined through a velocity and position control of an induction machine. It is shown that the fact that equality constraints could be considered, enables the use of tunings increasing the velocity of the structure, which could not be achieved with a classical GPC cascaded version.