An Experimental Assessment of Direct Torque Control and Model Predictive Control Methods for Induction Machine Drive

Finite-State Model Predictive Control methods (FSMPC) have been presented recently in the field of electrical drive and power electronics as an alternative to the conventional strategies. This paper presents a comparative evaluation between Direct Torque Control (DTC) and two finite-state model predictive control strategies applied to induction motor drive. Both DTC and MPC are nonlinear control techniques which dispense with the use of modulation unit (i.e. pulse width modulator (PWM) or space vector modulator (SVM)). DTC can provide good decoupled flux and torque control using pair of hysteresis comparators and lookup switching table for voltage vectors selection. In contrast with the model predictive control which includes the inverter model in control design. The optimal selection of inverter switching states minimizes the error between references and the predicted values of control variables by the optimization of a cost function. The effectiveness of applied algorithms is investigated by an experimental implementation using real-time interface (RTI) based on dSpace 1104.