Title: Comparison of static and dynamic characteristics of electromagnetic bearing using machine learning algorithm

Authors: Xiangxi Du; Yanhua Sun

Addresses: School of Mechanical Engineering, Xi'an Jiao Tong University, Xi'an, Shaanxi, China ' School of Mechanical Engineering, Xi'an Jiao Tong University, Xi'an, Shaanxi, China

Abstract: To improve the performance and service life of the bearing and improve the overall performance of the mechanical system, the characteristics of the electromagnetic bearing and the elastic foil gas bearing are analysed based on the machine learning algorithm. First, the bearing capacity of electromagnetic bearing is analysed, including the non-linear stiffness of electromagnetic bearing, the influence of air gap on the electromagnetic force, the determination of optimal linear range and the improvement of PID control based on support vector machine. At the same time, the characteristics of the elastic foil gas bearing are analysed, including the static characteristics of the bearing aeroelastic coupling calculation process and the calculation of dynamic stiffness and damping coefficient. The results show that with the gradual increase of the current, the radial electromagnetic force of the electromagnetic bearing also increases, and the increase range is larger and larger; when the current is constant, the electromagnetic force decreases with the increase of the air gap. When the frequency is small, the response curve of electromagnetic force of electromagnetic bearing fluctuates greatly with the change of control square wave. The research has practical reference value for the optimal design of electromagnetic bearing and elastic foil.

Keywords: aeroelastic coupling; machine learning; support vector machine; electromagnetic bearing; elastic foil gas bearing.

DOI: 10.1504/IJGUC.2022.121414

International Journal of Grid and Utility Computing, 2022 Vol.13 No.1, pp.87 - 95

Received: 04 Jan 2021
Accepted: 14 May 2021
Published online: 11 Mar 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article