2021/03/30 by Marouane Frini, Frini, Marouane, Vincent Choqueuse +3
Engineering · #Engineering Diagnostics and Reliability #FOS: Electrical engineering #Gear and Bearing Dynamics Analysis #Machine Fault Diagnosis Techniques #Non-Destructive Testing Techniques #Power Transformer Diagnostics and Insulation #Systems and Control (eess.SY) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2103.16165
openalex publication_date 2021/03/30 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
In the last few years, Motor Current Signature Analysis (MCSA) has proven to\nbe an effective method for electrical machines condition monitoring. Indeed,\nmany mechanical and electrical faults manifest as side-band spectral components\ngenerated around the fundamental frequency component of the motor current.\nThese components are called interharmonics and they are a major focus of fault\ndetection using MCSA. However, the main drawback of this approach is that the\ninterference of other more prevalent components can obstruct the effect of\ninterharmonics in the spectrum and may therefore impede fault detection\naccuracy. Thus, we propose in this paper an alternative approach that\ndecomposes the different current components based on the Vandermonde model and\nimplements the tracking of each distinct component in time and spectral\ndomains. This is achieved by estimating their respective relevant parameters\nusing the Gradient Descent algorithm. The results of this work prove to be\npromising and establish the parametric tracking of the electrical current\ncomponents using the Gradient Descent algorithm as a reliable monitoring\napproach.\n