Condition Monitoring of Induction Machines: Quantitative Analysis and Comparison

• Published in: Sensors (Basel, Switzerland) Vol. 23; no. 2; p. 1046
• Main Authors: Sintoni, Michele, Macrelli, Elena, Bellini, Alberto, Bianchini, Claudio
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 16.01.2023; MDPI
• Subjects: Algorithms; Asymmetry; Bandwidths; Computer Simulation; Diagnostic tests; electric machines; fault diagnosis; Methods; Quantitative analysis; Signal processing; Simulation; Spectrum allocation; Wavelet Analysis; Wavelet transforms; wavelet transforms; electric machines; fault diagnosis
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s23021046

In this paper, a diagnostic procedure for rotor bar faults in induction motors is presented, based on the Hilbert and discrete wavelet transforms. The method is compared with other procedures with the same data, which are based on time–frequency analysis, frequency analysis and time domain. The results show that this method improves the rotor fault detection in transient conditions. Variable speed drive applications are common in industry. However, traditional condition monitoring methods fail in time-varying conditions or with load oscillations. This method is based on the combined use of the Hilbert and discrete wavelet transforms, which compute the energy in a bandwidth corresponding to the maximum fault signature. Theoretical analysis, numerical simulation and experiments are presented, which confirm the enhanced performance of the proposed method with respect to prior solutions, especially in time-varying conditions. The comparison is based on quantitative analysis that helps in choosing the optimal trade-off between performance and (computational) cost.