Torque ripple reduction of induction motor based on a hybrid method of model predictive torque control and particle swarm optimization. (October 2016)

• Record Type: Journal Article
• Authors: Su, Te-Jen; Tsou, Tung-Yeh; Wang, Shih-Ming; Li, Tsung-Ying; Vu, Hong-Quan
• Is Part Of: Advances in mechanical engineering. Volume 8:Number 10(2016:Oct.)
• Publication Date: 2016-10
• Subjects: Model predictive torque control -- induction motor -- particle swarm optimization -- genetic algorithm -- space vector modulation -- torque ripple reduction; Mechanical engineering -- Periodicals; 621.05
• DOI: 10.1177/1687814016676465
• Languages: English
• ISSNs: 1687-8132

Abstract

The quality of torque of induction motors plays an important role in many electrical drive systems. Model predictive torque control has been a good alternative to conventional direct torque control for improving the torque qualities. Recently, some researchers tried to modify the model predictive torque control with other principles such as minimize torque ripple, optimize duty cycle, in an effort to get smaller torque ripples. However, according to the reported results the torque ripples are still significant. In this article, model predictive torque control based on particle swarm optimization is proposed to modify the model predictive torque control in order to improve the control qualities, especially the steady-state torque ripples. The key idea of this approach is using particle swarm optimization to minimize the cost function. The optimal voltage vector is implemented by space vector modulation technique. In addition, the control performance of model predictive torque control–particle swarm optimization combination is also compared with that of model predictive torque control–genetic algorithm to justify the effectiveness of our method. The presented simulations prove the better torque and phase currents quality at steady state of this approach.