Harmonic Current Optimization for Torque Ripple Reduction in Permanent Magnet Synchronous Machine Drives Based on Torque Ripple Surrogate Model

Torque ripple reduction based on harmonic current injection has been developed for PMSM drives. In such methods, torque ripple model (TRM) or speed harmonics are used for harmonic current optimization, which have several limitations. The performance of TRM based methods is limited due to the accuracy of the model itself and machine parameters, which leads to the remaining torque ripples. The speed harmonic based methods have speed limitations, since the speed harmonics generated by the torque ripples cannot be detected at high-speed operations. In this paper, the torque ripples of PMSM drives are first described based on the surrogate model, which does not require machine parameters. Based on that, the numerical solution of optimal harmonic currents is obtained offline using particle swarm optimization (PSO), considering both torque ripple reduction and loss minimization. Since the torque ripples are predicted with the machine-parameter-independent model instead of the speed harmonic measurements, the impact of the inaccuracy in machine parameters and the analytical torque ripple model are removed, and the proposed method may be effective over a broader range of speeds. The proposed method is evaluated experimentally and demonstrated to have several advantages over existing alternative methods in reducing torque ripples.