Comparison of spectral estimation techniques applied to induction motor broken bars detection

Abstract

This paper presents a performance comparison among some of the most effective spectral estimation techniques applied to the fault diagnosis o f induction machines. The diagnostic test is based on the analysis ofthe current space vector during motor starling via short-time analysis, using a sliding window and different spectral estimation algorithms. Differently from most o f the diagnostic techniques already proposed in the technical literature, the approach, presented in this work, i s effective regardless the load condition of the machine. Algorithms based on the FFT o r optimal band-pass filters (nonparametric methods), on the estimation o f a linear time-invariant model of the signal (parametric mefhods), and on the eigenanalysis of the autocorrelation matrix (high-resolution methods) have been used to process the motor current spare-vector. Experiments prove that both parametric and high-resolution methods overcame the FFT-based approaches, keep only the principal frequency components o f the signal and decrease the noise influence, thus permitting a better interpretation of the current vector spectrum and an automatic fault detection procedure.

A High-Resolution Frequency Estimation Method for Three-Phase Induction Machine Fault Detection

“…In all the works considering the application of this method for induction machine fault detection [14]- [16], the authors assumed a predefined value for this parameter without any justification. In this paper, an estimation method for the FSDO (m) value based on specification of the eigenvalues is applied.…”

“…These methods define a pseudospectrum function with large peaks that are subspace frequency estimates [13], and they are commonly used in the communication area. They have been recently introduced in the area of induction machine diagnosis by the application of the multiple signal classification (MUSIC) method [14]- [16]. This last method associates the noise to the minimum eigenvalues, which allows low SNR frequency components estimation.…”

“…With this last method, the frequency estimation is based on the orthogonality of the signal vectors to the noise eigenvectors. In [14], MUSIC is used to reduce the noise influence in the analyzed signal, whereas in [15], MUSIC allows the best frequency resolution in a dynamic signal. In [16], MUSIC is applied for a trivial case of the third and fifth harmonics components location without any explicit methodology.…”

“…These methods are classified in nonparametric, parametric, and subspace techniques [13]. In [14], the classical methods are applied to the stator current signal of an induction machine taking into account the fact that the signal is nonstationary. The comparison with other methods has shown that nonparametric methods have the same problems compared to the classical FFT concerning the frequency resolution.…”

Zoom-MUSIC frequency estimation method for three-phase induction machine fault detection

“…These methods are classified in nonparametric, parametric and subspace techniques [8]. In [9], the classical methods are applied to the stator current signal of the induction machine in non-stationary condition. The comparison with other methods has shown that nonparametric methods have the same problems compared to the classical FFT concerning the frequency resolution.…”