Modeling, simulation and implementation of a permanent magnet synchronous motor drive system using anfis technique

Abstract

The Permanent Magnet Synchronous Motors are widely used in robotics, electricity, gas, oil and transportation, etc. The PMSM drive allows adjustable speed drives and enhanced speed and torque control. Since the motor has various advantages such as higher efficiency, improved torque to inertia ratio and high power density, the PMSM drive has become very popular in aerial and electrical vehicles, as well as military applications [1-3]. Compared to the DC motors, the Permanent Magnet Synchronous Motors have distributed windings and do not have a brush-commutator system. They provide reliable operation. However, these motors require advanced control techniques to drive. Field Oriented Control (FOC) of PMSM is the most common method for the speed control of this motor. Employing Park and Clarke transformations, the control of the torque and flux of the motor is achieved by using appropriate field orientations which are dependent on phase currents in the stator. PMSM drive based on FOC combines the d-q-axis current control in the inner loop and the speed control in the outer loop[4]. The most significant requirements for the speed control of the PMSM are improved dynamic features and high-precision tracking performance [5]. The entire PMSM drive system is modelled using mathematical expressions of a PMSM and an inverter, the system is controlled with two different controllers and implemented through code generation in MATLAB/Simulink. The speed control with FOC method generally is integrated with a PI controller because of a simple design. As a drawback of a PI controller design for the PMSM, fine tuning of PI parameters is a challenging task since the linearization is utilized to converge the value of a function at a particular point. The parameter variations and disturbances cause undesired uncertainties in the non-linear model of the PMSM and the complexity is increased. The most of linear control methods cannot fulfill the requirements in high-performance applications of PMSM drives. To enhance the dynamic response of the PMSM control system, several advanced non-linear control methods have been studied and applied in academic and commercial fields recently, such as: linearization control [6], adaptive control [7,8], sliding mode control [9, 10], fuzzy control [11] and neural network control [12, 13].