Two-vector Dimensionless Model Predictive Control of PMSM Drives Based on Fuzzy Decision Making

Abstract

Model predictive controls (MPCs) with the merits of non-linear multi-variable control can achieve better performance than other commonly used control methods for permanent magnet synchronous motor (PMSM) drives. However, the conventional MPCs have various issues, including unsatisfactory steady-state performance, variable switching frequency, and difficult selection of appropriate weighting factors. This paper proposes two different improved MPC methods to deal with these issues. One method is the two-vector dimensionless model predictive torque control (MPTC). Two cost functions (torque and flux) and fuzzy decision-making are used to eliminate the weighting factor and select the first optimum vector. The torque cost function selects a second vector whose duty cycle is determined based on the torque error. The other method is the two-vector dimensionless model predictive current control (MPCC). The first vector is selected the same as in the conventional MPC method. Two separate current cost functions and fuzzy decision-making are used to select the second vector whose duty cycle is determined based on the current error. Both proposed methods utilize the space vector PWM modulator to regulate the switching frequency. Numerical simulation results show that the proposed methods have better steady-state and transient performances than the conventional MPCs and other existing improved MPCs.