A predictive optimal torque control of synchronous reluctance motor drive

Abstract

In this paper, a predictive digital control of the synchronous reluctance motor drive is presented. The proposed control, implemented in rotor coordinates, is based on a recursive algorithm in which the reference values of d and q current components are imposed. The digital control system directly synthesizes the output vector which allows the timing control of the inverter by means of on/off state of the switches. This results in avoiding the implementation of the traditional PI controllers and PWM technique. Under transient operation, both the reference d-q current components are set by imposing the maximum allowable torque variation, during the actual sampling interval, taking into account the inverter voltage saturation. In this way, the optimal (fastest) torque trajectory can be followed. The proposed algorithm guarantees optimal torque step response. The d-q current values, corresponding to steady state torque, do not guarantee any other additional condition. Hence, after reference torque is reached, the drive can be controlled by means of an additional predictive algorithm which assures minimum Joule losses. A computer simulation study, using Matlab Simulink tool, is conducted on the drive with the aim of highlighting the good performance of the proposed algorithms compared to traditional and to predictive (not optimal) ones.