Modified LMS algorithm applied to maglev flywheel's vibration compensation problem

Article type: Research Article

Authors: Gao, Hui^* | Song, Le

Affiliations: School of Automation, Nanjing University of Posts and Telecommunications, Nanjing, Jiangsu, China

Correspondence: [*] Corresponding author: Hui Gao, School of Automation, Nanjing University of Posts and Telecommunications, Nanjing 210023, Jiangsu, China. E-mail:gaoh@njupt.edu.cn

Abstract: For the high-speed rotation characteristics of maglev flywheel, in order to reduce the vibration impact of shared frequency sinusoidal signal with the flywheel, an improved variable step size least mean square (LMS) algorithm is proposed through analyzing the application deficiency in maglev flywheel system of the standard LMS algorithm and original variable step size LMS algorithm. The filtering compensation effect is compared of the proposed LMS algorithm with the original algorithm. Simulation results show that the modified algorithm having better convergence and follow the desired signal ability to time-varying small amplitude signals. The study can offer help to improve the control stability and compensation real-time performance for maglev flywheel system.

Keywords: Maglev flywheel, frequency mismatch, variable step-size LMS algorithm, frequency time-varying, adaptive filtering

DOI: 10.3233/JAE-160068

Journal: International Journal of Applied Electromagnetics and Mechanics, vol. 53, no. 2, pp. 359-369, 2017