Diagnostics of Low-power Electric Machines Based on the Analysis of Energy Spectrum of Signals

The aim of the article is to develop a new digital method for diagnosing electrical machines using the energy spectrum of electrical signals based on parametric discrete Fourier transform. It is shown that the development of new and improvement of existing methods for diagnosing electric machines is an important and urgent task of increasing the efficiency of operation and the reliability of the operation of equipment in various industries. The classification of methods of functional diagnostics by the actual state into the following groups is given: methods for analyzing a vibro-acoustic signal, magnetic flux, secondary electromagnetic fields, temperature fields, electrical parameters of electrical machines during their operation. It is shown that the most promising group of methods for the functional diagnosis of electrical machines at nuclear power plants are methods for analyzing the electrical parameters of electrical machines during their operation. The physical basis and advantages of this group of methods for the functional diagnosis of electrical machines are given. It is shown that the digital technology implementation, the launch of the digitalization process for diagnosing objects, the transition in the diagnosis of objects from analog signal processing (ASP) to digital signal processing (DSP) methods has led to a number of theoretical and practical problems. The paper considers discrete Fourier transform (DFT) (algebraic form), generalization of DFT in the form of parametric DFT (DFT-P) (matrix form), and the analytic properties of their bases. Methods for the fast calculation of DFT and DFT-P are given. The question of choosing a strategy for digital spectral analysis of mixed signals is considered from the point of view of determining the parameters of the harmonic components of the signals of a functioning electric machine, since these parameters are really informative. An experiment conducted by numerical simulation methods confirmed the theoretical results obtained in the article.