Online fuzzy tuning of weighting factor in model predictive control of PMSM

Abstract

Permanent magnet synchronous motors (PMSMs) have sensible characteristics which make them attractive in industrial applications such as high power density, high efficiency and high torque. Hence they are suitable for a wide variety of applications, especially where weight and size are restricted. Several methods have been proposed to control PMSMs such as field oriented control (FOC) and direct torque control (DTC). Utilization of the PWM and a PI as a current controller is the common feature in the mentioned methods, while the model predictive control (MPC) generates the switching signals directly by optimization of a cost function and as will be explained, doesn't use a PI to control the motor currents. In this paper a fuzzy procedure is applied to tune the weighting factor of the cost function online as the reference changes. Because the weighting factor affects the torque ripple and currents ripple and should be changed to obtain the best performance during the variation of reference. A program in matlab has been developed to verify the efficiency of the method. The results are demonstrated and compared in the following sections.