A new approach to reduce torque ripple and noise in twelve-sector based direct torque controller fed permanent magnet synchronous motor drive: simulation and experimental results

Abstract

The classical direct torque control (DTC) provides large torque ripple. Several techniques have been attempted on torque ripple reduction, in which most of them are usually complicated and have more computational burden. In this study, the twelve-sector based DTC for permanent magnet synchronous motor (PMSM) with reduced ripples in the torque is presented. First, the structure of the classical DTC is modified and which provides 27 voltage vectors with different magnitudes, namely, Large (L), Medium (M), Small (S), and Zero (Z). This study also presents six new switching tables by arranging the voltage vectors effectively, such as Largeâ–“Zero (LZ), Mediumâ–“Zero (MZ), Smallâ–“Zero (SZ), Largeâ–“Mediumâ–“Zero (LMZ), Largeâ–“Smallâ–“Zero (LSZ), Mediumâ–“Smallâ–“Zero (MSZ). The proposed switching table based DTC reduces the torque ripple very effectively. Moreover, this method only modifies the switching table and, hence, greatly reduces the computational burden. To improve the system performance and to reduce the torque ripple further, the adaptive neuro-fuzzy inference system (ANFIS) based controllers are used to adjust the gain in the controllers. The improved performances are obtained in terms of torque ripple, mechanical vibration and noise. Finally, the experimental studies are presented to validate the simulation results.