Induction motor bearing fault analysis using Root-MUSIC method

Abstract

This paper describes a new diagnosis approach, the Root-MUSIC (RM) method, for identification of the progressive cracking in the bearing of induction motors. This approach has several advantages compared to the stator current spectral analysis using the conventional Periodogram method. Indeed, the main advantage of this approach is its very good frequency resolution for a very short acquisition time, something impossible to achieve with the conventional method. However, in order to reduce the computation time which is the main drawback of the RM method, this method will be applied to only a specified frequency band; one that carries information about the sought fault. Experimental results show the effectiveness of RM method on the reliability of the incipient bearing fault detection.

Bearing Corrosion Failure Diagnosis of Doubly Fed Induction Generator in Wind Turbines Based on Stator Current Analysis

“…Considering the energy of such signals, the power spectrum can be introduced to evaluate the power within the unit frequency band [16]. Several improved methods have been proposed to increase the accuracy and frequency resolution of power spectra, e.g., the spectral kurtosis algorithm [10], the Welch power spectrum analysis [17], the root-MUSIC method [18] and noise cancellation [19]. However, the Fourier transform cannot indicate how the frequency content of a non-stationary signal changes.…”

Use of a combined SVD-Kalman filter approach for induction motor broken rotor bars identification

“…For example, the MUltiple SIgnal Classification (MUSIC) method [13] and its Root-MUSIC variant are based on the noise subspace, and the Estimation of Signal Parameters using Rotational Invariance Techniques (ESPRIT) method [14] is based on the signal subspace. The applications of these methods for the electrical machine faults diagnosis are detailed in [15], [16]. It should be noted finally, that the principal limitations of these methods are due to their low robustness to the unknown signal components number [15] and the unreliable estimation of the amplitudes and phases of the sought harmonics [17].…”

“…The applications of these methods for the electrical machine faults diagnosis are detailed in [15], [16]. It should be noted finally, that the principal limitations of these methods are due to their low robustness to the unknown signal components number [15] and the unreliable estimation of the amplitudes and phases of the sought harmonics [17]. In addition, these methods require a significant computation time.…”

“…Nevertheless, this approach requires an important computation time, which can be a serious handicap for its implementation in real time. However, since the frequency signature of a fault is located on a known frequency band [15], then the proposed solution in this paper consists of the decrease of the computation time and the used memory space by applying the processing only on the frequency band where the signature of the fault is likely to appear. To highlight the contribution merits of this new approach in the detection and the monitoring of rotor faults severity, several experimental tests are conducted on an induction motor by analysing the stator current signal in the steady-state and without varying the load or speed of the motor.…”

“…The suggested idea consists, therefore, to apply the proposed technique SVD-Kalman Filter only on this frequency band and not on the entire spectrum of the stator current. This solution will significantly reduce the length of the signal to be analysed and therefore the computation time[15].With this solution, the processing will thus be performed on the N starting samples. N b is the number of samples with…”