Predictive control of PMSM using improved model of the drive

Abstract

The most common control algorithms for PMSM are based on simplified interpretation of the mathematical model, assuming strictly harmonic signals and linear distribution of the magnetic flux of permanent magnets in the air gap. Nevertheless neglected phenomenons in the mathematical mode leads to deterioration of the torque and overall it may lower the efficiency of the machine. Therefore in order to improve efficiency and torque quality, the aforementioned effects needs to be taken into account, when designing mathematical model, where, stator inductance and magnetic flux of permanent magnets are considered a function of the rotor position. the optimal reference current is no longer projected into d, q coordinated system as a static dc value, which significantly increase the demands on the tuning of cascade structure of PI controllers. An alternative is to use predictive control.In this paper we seek for simple and yet robust control with comparable results as a standard solution based on d, q coordinated system and linear controllers. For this purpose we use simple version of predictive control working with limited control action set (FCS-MPC) with precise model defined in rotating d, q reference frame The resulting algorithm provides high quality torque control performance while taking into account Joule losses in the motor.