Predictive DTC schemes with PI regulator and particle swarm optimization for PMSM drive: comparative simulation and experimental study

Abstract

This paper shows that a predictive control (PC) combined with the principle of direct torque control (DTC) leads to an excellent dynamic behavior of the permanent magnets synchronous machine (PMSM) and mitigates the drawbacks of the conventional DTC. In this paper, a predictive direct torque control (PDTC) of a PMSM based on classical PI regulator and particle swarm optimization (PSO) algorithm is developed. The PSO adjusts the parameters of the PI controller, which improves its adjustive capability. The effectiveness of the proposed PDTC based a PSO controller is then compared with that of the conventional PI controller. The numerical simulation is performed using MATLAB-Simulink and SimPowerSystem Toolbox. The main contribution of this work is the implementation of the proposed controller on a test bench around a dSPACE 1104. Simulation and experimental results are presented in order to demonstrate the effectiveness of the proposed techniques. Besides, the system associated with these techniques reduces effectively the flux and the torque ripples with better dynamic and steady state performance. The results with the PSO show more performances and a considerable reduction in torque and flux ripples than the results with the conventional PI.