The seminar talk mainly covers formulation of simplified model predictive current control (MPCC) methods for a three-level open-end winding induction motor (OEWIM) drive that are mainly aimed at reduced computational burden on the digital controller. While dealing with the three-level OEWIM drive during transition between two zero CMV switching states, effectively a non-zero CMV appears during the dead time interval depending on the position of the stator current vector.

In the first part of the talk, a simple MPCC method for complete elimination of CMV shall be discussed specially suited in low-speed operation. It is found that the constraint on eliminating the dead time effect on CMV heavily deteriorates the performance of the drive at higher speeds. Hence, a smooth transition with and without the constraint shall also be discussed.

In the second part of the talk, the OEWIM drive is operated with a single dc source on one side and a floating capacitor bank on th eother side. For this arrangement, a cascaded predictive current control (CPCC) method is proposed that completely avoids the use of weighting factors. Sequential control of multi-objectives by identifying the order of evaluation of individual objectives is introduced. The motor torque and flux are controlled first followed by voltage balancing of floating capacitor in order to determine the optimal switching combination of the 3-level inverter drive.

In the present work, the traditional current based objective function is replaced with an equivalent voltage based objective function in order to reduce the computational burden in both the envisaged MPC methods. Ruggedness of these MPC methods shall be demonstrated through simulations and experimental studies covering speeds below and above motor base speeds of operation. Performance of the proposed MPC and the existing MPC methods shall also be presented to showcase the merits of the proposed MPC methods. Above all, owing to the simplicity of the proposed MPC, a drastic reduction in the execution times achieved when compared to the existing MPC methods shall be presented.