Design of Neuro-Fuzzy Based Torque Controller for Torque Ripple Reduction of Asynchronous Motor

Authors : Ahmet GÜNDOĞDU, Fikret ATA, Beşir DANDIL

Pages : 225-234

Doi:10.17694/bajece.730311

View : 23 | Download : 7

Publication Date : 2020-07-30

Article Type : Research Paper

Abstract :Among all the control methods developed for Induction Motor insert ignore into journalissuearticles values(IM); drivers, the hysteresis controller based Direct Torque Control insert ignore into journalissuearticles values(DTC); method has an important place. This control method does not require other rotor and stator parameters except the stator resistance and does not require position or velocity sensors. However, there are some disadvantages of the DTC method, such as high torque, flux and current ripples. In this study, in order to reduce the high torque ripples occurring in an induction motor that is controlled by the hysteresis controller based conventional DTC method, a simple and effective Sugeno type Neuro-Fuzzy Torque Controller insert ignore into journalissuearticles values(NFTC); is proposed. This proposed controller is used instead of hysteresis controller. An experimental setup consisting of 1.1 kW induction motor, current and voltage measurement, DS1103 control card and two-level voltage source inverter was installed. To evaluate the performance of the proposed controller structure, various experimental studies were performed. Results obtained from the proposed NFTC based structure and conventional hysteresis controller based DTC structures are given comparatively. By the obtained experimental results, it was confirmed that the proposed NFTC-based controller structure considerably reduced flux and torque ripples in the motor.
Keywords : Direct Torque Control, AC Drives, Neural Fuzzy Networks, Torque Ripple Reduction