Performance Evaluation of PMSM Drive Using ANFIS based Speed Controller Under SVPWM Technique

Abstract

In this paper, a adaptive neuro-fuzzy inference system (ANFIS) based speed controller is presented for permanent magnet synchronous motor (PMSM). In high-performance applications, AC drives utilize space vector pulse width modulation(SVPWM) technique to operate like separately excited DC motors. However, the fuzzy and artificial neural network based speed controller for PMSM has their own demerits, that cannot be avoided. But, combined effect of neuro-fuzzy controller can conquer their drawbacks and effectually controls the speed. For comparative analysis, proportional-integrator and ANFIS controller are employed for the speed loop and a SVPWM technique is utilized in current loop for the impressive torque control. In addition, the paper includes an overview of PMSM modeling, SVPWM scheme and ANFIS controller. The MATLAB simulations are conducted to assess the performance of the PMSM drive. In simulations, starting current, torque ripples, harmonic contain in stator current, and settling time required for speed controllers are compared. Also, hardware prototype for PMSM drive is developed in the laboratory and tested partially for the SVPWM scheme.