Ananfis based torque ripple minimization in permanent magnet stepper motor

Abstract

In this research work an efficient technique is used for decreasing newlinethe torque ripple of the PMSM. This is on the basis of the hybridization of newlineANN and fuzzy inference system that is called as ANFIS technique, which is newlinehighly efficient in nonlinear systems because of the fact that once properly newlinetrained they can interpolate and extrapolate the random information with high newlineaccuracy. The main objective of the projected method is to reduce the torque newlineripples with the help of the regulating parameters, namely torque and speed. newlineInitially the PMSM parameters are restrained and controlling the input newlineparameters of the PMSM such as voltage and current. The torque and rotor newlineangle of the PMSM is restrained for controlling the torque ripple and newlineregulating the speed of the PMSM. From the measured parameters the error newlinesignal is considered from the actual and reference value of the rotor angle and newlinetorque of the PMSM. On the basis of the error signal the projected method is newlineproduced the control pulses, which is provided to VSC for supply the voltage newlineand current signal to PMSM. The presented torque ripple minimization newlinemethod is applied in MATLAB/Simulink working platform and the newlineperformances is assessed and compared with some available methods such as newlineneural networks and fuzzy controller. newlineThe speed of the Stepper motor can be adjusted to a great extent so newlineas to provide easy control and high performance. In this research an effort is newlinemade to develop a prototype model to control the speed of a stepper motor newlineusing PSO-ANFIS hybrid technique. We describe the use of Particle Swarm newlineOptimization (PSO) and ANFIS for designing an optimal fuzzy logic newlinecontroller of a stepper Motor. We have used two input parameters like speed, newlinetorque and output is step angle. newline