Analysis and Optimization of Torque Ripple Reduction Strategy of Surface-Mounted Permanent-Magnet Motors in Flux-Weakening Region Based on Genetic Algorithm

One of the parameters required for a high-performance drive of a surface-mounted permanent magnet synchronous motor (SPMSM) operating in the flux-weakening region is the torque ripple reduction. In general, the torque ripple results from many reasons, such as the machine structure, oscillations of measured speed, etc. Hence, this article proposes an optimized strategy that works in conjunction with a variable switching frequency pulsewidth modulation algorithm to reduce the torque ripple peak. For the modulating algorithm, the updated frequency changes linearly with the desired torque ripple and its predicted maximum peak. For the proposed strategy, the reflection of proportional-integral (PI) parameters is observed at the torque ripple caused by the current ripples. Therefore, a robust method for adjusting PI parameters is proposed that relies on the improved fitness function that can minimize the regulators' error and maximize the drive stability bandwidth in the flux-weakening region. This objective function is optimized offline using a genetic algorithm optimization technique. Meanwhile, a third-order generalized integral flux observer is applied in this article, providing evidence of torque ripple reduction. For driving in the flux-weakening region, the reference magnitude of the duty cycles can produce the required d-axis reference current to prevent the saturation of current controllers. Finally, comprehensive simulations and experiments are presented to validate the proposed strategy effectively. In comparison with the conventional method, the proposed strategy can reduce the copper and switching losses simultaneously and control the current and speed in the flux-weakening region efficiently.