Reducing torque ripple for switched reluctance motors by current reshaping neural network

Abstract The high torque ripple poses a limitation on the application of switched reluctance motors (SRMs). In this paper, a current reshaping neural network (CRNN) is proposed to mitigate the torque ripple. The principle of how current affects electromagnetic torque is analyzed bas the indirect tor...

Full description

Saved in:
Bibliographic Details
Main Authors: Benqin Jing, Guofu Liang, Xuanju Dang, Yanjun Jiang
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-02228-z
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Abstract The high torque ripple poses a limitation on the application of switched reluctance motors (SRMs). In this paper, a current reshaping neural network (CRNN) is proposed to mitigate the torque ripple. The principle of how current affects electromagnetic torque is analyzed bas the indirect torque control method. In order to reduce the torque ripple, the CRNN is introduced to establish a current model and generate precise current by connecting the total current and the current sharing section. Additionally, the CRNN’s implicit function is constructed based on the total current and rotor angle, with the weight being adjusted through proportion differentiation compensation current. Ultimately, the CRNN effectively reduces the torque ripple by modifying the phase current profiling. The efficacy of the proposed method is validated through comprehensive simulations and experiments on a three-phase 12/8 SRMs, conducted under various operating conditions.
ISSN:2045-2322