TORQUE RIPPLE MINIMIZATION IN PERMANENT MAGNET SYNCHRONOUS MOTOR DRIVE

Abstract

Use of permanent magnets made of rare earth materials such as samarium cobalt and neodymium-boron-iron in Permanent Magnet Synchronous Motor (PMSM) drives has resulted in high flux density and improved performance of the drive. Field Oriented Control (FOC) has become one of the most popular speed and torque control techniques for AC motors. In PMSM drive detection/computation of rotor position is crucial for ensuring high performance during FOC. Rotor position is often sensed by incremental encoders or resolvers. The use of positon sensors in motor speed control increases the cost, size, weight and wiring complexity, and reduces the mechanical robustness and reliability of the PMSM drive systems. Sensor-less speed control techniques overcome these drawbacks related to estimation of speed and rotor position. The PMSMs are generally employed in industrial servo applications because of their fast dynamic performance. However, PMSMs suffers from ripples in the torque produced. Torque ripples in PMSM are produced because of cogging, current measurement error, switching of inverter and harmonics in magnetic flux. Torque ripples also leads to fluctuations in speed thus limiting the use of PMSM in several servo applications. Torque ripples could be minimized in applications that demand accurate speed/position tracking. The present work aims to explore use of different modern control techniques to minimize torque ripples in the operation of PMSM drives in comparison to previously reported control techniques. The objectives of the research include – a) modelling, design and development of laboratory prototype of PMSM drive, b) design and implementation of improved artificial neuro fuzzy inference system (ANFIS) based model reference adaptive control (MRAC) observer for sensor-less control of PMSM,

v

c) minimization of stator current ripples and torque ripples in PMSM drive using advanced predictive current controller (APCC) based on Dead Beat (DB) control theory d) minimization of torque ripples using intelligent hybrid controller (IHC) and

e) torque ripple minimization by model predictive control of PMSM using proportional-plus- integral resonant (PI-RES) controller.

The strategies to reduce torque ripples, that have been reported in the literature, may be classified into a) approaches based on the design improvement of the motor, b) methods based on control techniques or c) a combination of these two. The most critical aspect in high performance drive is the choice of the control algorithm that minimizes the torque ripples effectively for a given application. In this research work, a laboratory hardware prototype is designed and developed for real time analysis of PMSM drive based on sensor-less field-oriented control. An experimental setup is developed for implementation of FOC on PMSM using dSPACE DS1104 controller and performance of the drive is analysed using different control techniques. An improved ANFIS based MRAC observer is designed and implemented for FOC of PMSM with space vector PWM (SVPWM). In the proposed method adaptive model and adaptive mechanism are replaced by an improved ANFIS controller, which neutralize the effect of parametric variation and results in improved performance of the drive. The required rotor position and speed are estimated using the proposed MRAC observer. Simulation studies using MATLAB/Simulink and comparative analysis of the conventional MRAC based observer with improved ANFIS based MRAC observer show that better dynamic performance of the PMSM drive is achieved using the improved ANFIS based MRAC.

vi

An advanced predictive controller (APCC) based on deadbeat (DB) control theory is also developed and analysed for reduction of torque ripples in PMSM. The performance of the proposed APCC based on DB control theory are compared with hysteresis based direct current controller (DCC) and duty cycle-based model predictive controller (Duty-MPCC) under different operating condition through simulation studies using MATLAB/Simulink. It is observed that the implementation of proposed APCC results in better dynamic performance with less ripples in torque and stator currents, and lesser THD in stator current as compared to DCC and Duty-MPCC. An intelligent hybrid controller (IHC) has also been developed and implemented for FOC of PMSM to minimize torque ripples for constant torque operation. The proposed IHC is designed by combining a fuzzy logic controller (FLC) with PI controller with a novel switching capability. The intelligent switching decision of the developed IHC is based on overshoots, undershoots and oscillations observed in the system. Simulation studies for the FOC of PMSM using the proposed IHC indicates better dynamic performance with lesser torque ripples and lower THD in stator current in comparison with conventional PI controller. In addition, a proportional-plus-integral resonant (PI-RES) controller is designed and implemented for FOC of PMSM with model predictive controller (MPC) to minimize torque ripples for constant torque operation. The MPC is designed to provide the optimal voltage vector by minimizing the objective function calculated from stator current prediction for k th instant. The PI-RES controller is developed by combining a resonant controller with PI controller. Due to the compensating torque current produced by the resonant controller and

reference current from the PI controller, ripples in the speed response are minimized. A PI- RES controller generates the reference pulsating torque current, which counteracts the ripples

vii

in load torque. The FOC of PMSM with MPC using PI-RES is simulated in MATAB/Simulink and the performance of the drive is compared with MPC using PI controller. The proposed FOC of PMSM with MPC using PI-RES demonstrate better dynamic performance, lower torque ripples and lower THD in stator current in comparison with conventional PI controller based MPC.