A Wavelet-Based Analysis for Monitoring Controller Reliability in Active Magnetic Bearing With Rotor Eccentricities

Debarghya Dutta, Pabitra Kumar Biswas, Sukanta Debnath, Furkan Ahmad*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

The complexity of improving controllers for Active Magnetic Bearing Systems (AMBs), which are essential parts of fast electronic transport networks such as electric cars, aviation technology, and defense systems is examined in this research work. It evaluates AMBs' robustness against abrupt harmonic disruptions using wavelet transform methodologies. Initially, predictive-based interpolation equations are used to create system controller matrices that forecast controlling gains for various eccentric rotor behaviors. Steady functioning is ensured by the non-linear AMB system's assumed linearity inside this range. Furthermore, system dynamics are assessed about probable external defects by creating a modeled harmonic signal and using wavelet transformations in continuous and discrete realms. The reliability of the assessment method is highlighted by the low 1.15% variance across the original and recovered force signals at high rotor eccentricity, which is supported by suitable scalability and wavelet selection based on Fisher's criterion. The computational findings provide insights and enhance the understanding of regulating the fluctuating actions of AMBs through these thorough investigations, stimulating improved reliability and effectiveness in an array of applications.

Original languageEnglish
Pages (from-to)197335-197346
Number of pages12
JournalIEEE Access
Volume12
DOIs
Publication statusPublished - 11 Dec 2024

Keywords

  • Active magnetic bearing
  • control theory
  • controller
  • signal processing
  • vibration
  • wavelet transform

Fingerprint

Dive into the research topics of 'A Wavelet-Based Analysis for Monitoring Controller Reliability in Active Magnetic Bearing With Rotor Eccentricities'. Together they form a unique fingerprint.

Cite this